Intelligent underwater dynamic oil containment boom traction system based on multi-source cooperation and use method

Abstract

The application discloses a multi-source cooperative intelligent underwater power oil fence traction system, which comprises a main controller, an environment sensing unit, a navigation and positioning unit, a self-adaptive adjustment unit, a dynamic modeling and intelligent decision unit, a redundant system unit, an intelligent control unit, a power management unit and multiple underwater power devices, the main controller is responsible for overall task allocation and coordination, the environment sensing unit is used for identifying and tracking the oil fence, the navigation and positioning unit realizes multi-source sound and light fusion positioning, the redundant system unit is used for function takeover when part of the devices fails, the intelligent control unit monitors the state of each underwater power device in real time, and the power management unit monitors the power state of each underwater power device in real time. The multiple underwater power devices cooperate to provide strong traction, can effectively cope with the heavy oil fence laying and dragging demand under the condition of strong wind and wave, and significantly improve the oil spill recovery efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ocean engineering, in particular to an intelligent underwater dynamic oil containment boom traction system based on multi-source collaboration and a use method thereof. BACKGROUND

[0002] Marine oil spill accidents have caused serious impact on the environment and economy, therefore, quickly and effectively recovering the spilled oil has become an important research topic. The traditional oil spill recovery method mainly relies on mechanical recovery, spraying oil spill dispersants and controllable combustion technology. However, these methods have many technical defects and limitations when facing large-area oil spill and complex sea conditions.

[0003] In addition, there are unmanned automatic towing devices, but the traction force is insufficient: the existing unmanned automatic towing devices (such as the automatic navigation and automatic recovery electric oil containment boom and its operation method disclosed in application number CN201910740094.9) can realize unmanned operation, but still have limitations in traction force, and are difficult to meet the deployment and towing requirements of heavy oil containment booms.

[0004] Meanwhile, there are the following problems:

[0005] The traction force of a single unmanned device is insufficient, and multiple devices need to work collaboratively. However, the existing technology has technical problems in multi-device collaborative control and error correction, and it is difficult to achieve efficient collaborative work.

[0006] The data fusion accuracy is low, and the existing multi-source data fusion method has low accuracy in complex sea conditions, which is difficult to provide high-precision state estimation and error correction, affecting the overall performance of the system.

[0007] The real-time performance is poor, and the traditional data fusion method is complex and has poor real-time performance, which is difficult to meet the real-time control requirements of marine oil spill recovery operations.

[0008] In summary, the existing technology has technical defects such as insufficient traction force, complex operation, poor environmental adaptability, environmental pollution, high safety risk, collaborative work problems and low data fusion accuracy in marine oil spill recovery operations, and there is an urgent need for a new intelligent underwater dynamic oil containment boom traction system that can effectively solve the above problems and improve the efficiency and safety of oil spill recovery. SUMMARY

[0009] The technical problem to be solved by the present application is the technical defect of insufficient traction force, complex operation, poor environmental adaptability, environmental pollution, high safety risk, collaborative work problems and low data fusion accuracy in the existing marine oil spill recovery operations.

[0010] In order to solve the above technical problems, the technical scheme of the present application is to provide a multi-source collaborative-based intelligent underwater power oil containment boom traction system, which comprises a main controller, an environment perception unit, a navigation and positioning unit, an adaptive adjustment unit, a dynamic modeling and intelligent decision unit, a redundant system unit, an intelligent control unit, a power management unit and a plurality of underwater power devices, wherein,

[0011] The underwater power device is connected with the power management unit and comprises a propeller, a control module, a power module and a communication module, and each underwater power device is equipped with a local controller with autonomous decision-making capability;

[0012] The main controller is connected with the environment perception unit and is responsible for overall task allocation and coordination to ensure that the plurality of underwater power devices work collaboratively according to a predetermined strategy;

[0013] The environment perception unit is connected with the adaptive adjustment unit and the navigation and positioning unit and is integrated with a sonar, a camera and a current meter for identifying and tracking the oil containment boom;

[0014] The navigation and positioning unit comprises an acoustic sensor and an optical sensor to realize multi-source acoustic-optical fusion positioning, wherein the acoustic sensor is used for long-distance positioning and the optical sensor is used for short-distance accurate positioning;

[0015] The adaptive adjustment unit is connected with the dynamic modeling and intelligent decision unit and comprises a decision system based on fuzzy logic or neural network and a data fusion system based on Kalman filtering or multi-sensor data fusion algorithm;

[0016] The dynamic modeling and intelligent decision unit is connected with the redundant system unit and the intelligent control unit and comprises a deep learning model and a reinforcement learning intelligent agent, which uses a system identification model to update model parameters in real time;

[0017] The redundant system unit is used for function takeover when part of the device fails;

[0018] The intelligent control unit is connected with the power management unit, monitors the state of each underwater power device in real time, adjusts the working state according to the actual situation, realizes collaborative control and error correction of the multi-source underwater power device, and realizes collaborative control and error correction of the multi-source underwater power device;

[0019] The power management unit monitors the power state of each underwater power device in real time.

[0020] Optionally, the main controller comprises a collaborative control strategy module, the collaborative control strategy module adopts a consensus algorithm for realizing information exchange and collaborative action among the plurality of underwater power devices, and the real-time data provided by the environment perception module is used as the input of the consensus algorithm.

[0021] Optionally, the redundant system unit is equipped with a backup controller, which is connected to the main controller through a high-speed communication link.

[0022] Optionally, the control module in the underwater power device is equipped with a microprocessor, and is equipped with sensors for obtaining the status of the underwater power device. The power module uses replaceable lithium batteries, and the communication module is used for communication with other devices inside the system and the external control center.

[0023] Optionally, it also includes:

[0024] A data storage module for storing system operation data and sensor data;

[0025] A fault diagnosis module for detecting system faults and providing recommended solutions;

[0026] A simulation training module for simulating different operating environments and fault conditions to train the system response:

[0027] A remote control module that allows operators to remotely monitor and control system operations;

[0028] Modular design and standardized interface for quick assembly, maintenance and upgrade of the system;

[0029] Software configuration system using international standard-based communication and control interface.

[0030] The use method of the intelligent underwater power boom traction system based on multi-source cooperation includes the following steps:

[0031] S1, system initialization, the main controller assigns specific tasks to each underwater power device according to the preset task plan, including initial position, path planning and working mode;

[0032] S2, during the boom inflation and deployment process, the operator or automatic equipment installs underwater power devices on the boom at predetermined intervals;

[0033] S3, each underwater power device adjusts the thrust and direction of the propeller according to the instructions of the main controller, and cooperatively provides sufficient traction to make the boom work in the predetermined area;

[0034] S4, the power management unit monitors the power state and state of charge of each underwater power device in real time through voltage, current and temperature sensors;

[0035] S5, the redundant system unit monitors the status of each underwater power device in real time, and when a device failure is detected, immediately takes over its functions;

[0036] S6, the dynamic modeling and intelligent decision unit utilizes deep learning and reinforcement learning techniques to update model parameters in real time and optimize system performance;

[0037] S7, after the completion of the task, the main controller issues a recycling instruction, and each underwater power device stops working and returns to the designated position;

[0038] Optionally, in step S4, when the power is lower than the set threshold, the power management unit starts the charging process, controls the charging device to perform constant current charging, constant voltage charging or pulse charging, ensures that the battery is charged within a safe range, and when the state of charge of the battery is lower than the set threshold, the power management unit prompts that the battery needs to be replaced, and coordinates the battery replacement process.

[0039] Optionally, step S5 includes:

[0040] S51, the intelligent control unit analyzes the sensor data on the underwater power device, identifies abnormal conditions, and when an abnormal condition is detected, the system issues a fault alarm and records the fault information;

[0041] S52, the main controller and the backup controller synchronize data in real time to ensure that the backup controller can take over at any time, and when the main controller fails, the backup controller automatically takes over its functions;

[0042] S53, real-time detection of system failure through sensors and monitoring systems, isolation of failed components, and prevention of fault propagation;

[0043] S54, the redundant system unit takes over the functions of the failed components to ensure the reliability and continuity of the system, and the system records fault information and prompts the operator to repair the fault;

[0044] S55, after repair, the system is reinitialized and returns to normal operation, and the redundant system unit synchronizes the data during the fault to the main controller.

[0045] Optionally, step S6 includes:

[0046] S61, collect sensor data during system operation, and collect control signals during system operation;

[0047] S62, clean, normalize and segment the collected data to prepare training and test data sets;

[0048] S63, design the structure of the deep neural network, including the input layer, multiple hidden layers and the output layer, use the training data set, adjust the network weight and bias through the back propagation algorithm, so that the model can accurately predict the state and behavior of the system;

[0049] S64, verify the performance of the model using the test data set, adjust the hyperparameters to optimize the model, deploy the trained model into the system to predict and adjust the system state in real time;

[0050] S65, define the state space, action space and reward function of the system, build a simulation environment, design a reinforcement learning agent including a policy network and a value network;

[0051] S66, make the agent explore and learn in the simulation environment, and continuously adjust the policy through trial and error to maximize the cumulative reward;

[0052] S67, use optimization algorithms such as gradient descent to adjust the parameters of the policy network to improve the performance of the policy, and deploy the trained policy into the system to make real-time decisions and control system behavior;

[0053] S68, collect the input and output data of the system, including sensor data and control signals. Select a suitable model structure, estimate the model parameters, so that the model can accurately describe the dynamic behavior of the system, use independent data sets to verify the performance of the model, adjust the model structure and parameters to optimize the model, and update the model parameters in real time during the operation of the system to adapt to the dynamic changes of the system.

[0054] Optionally, step S7 comprises:

[0055] S71, the main controller sends a recovery instruction to each underwater power device according to the task completion, and each underwater power device stops the current work after receiving the recovery instruction and prepares to return to the specified position;

[0056] S72, the navigation and positioning unit plans the optimal recovery path according to the current environment and the position of each underwater power device, and the self-adaptive adjustment unit dynamically adjusts the recovery path of each underwater power device according to real-time environmental data and task requirements;

[0057] S73, each underwater power device coordinates the return through a consistency algorithm to avoid conflict and repeated work, and the intelligent control unit monitors the state of each underwater power device in real time;

[0058] S74, after each underwater power device returns to the specified position, personnel or automatic equipment recovers and maintains the underwater power device, the power management unit checks the power state of each underwater power device and performs necessary charging or battery replacement, and the system records the data during the recovery process for analysis and optimization to prepare for the next use.

[0059] In summary, the present application has at least one of the following beneficial effects:

[0060] 1. The present invention provides powerful traction through the coordinated work of multiple underwater power devices, which can effectively cope with the needs of deploying and towing heavy oil booms in strong winds and waves, and significantly improve the efficiency of oil spill recovery.

[0061] 2. The present invention adopts an improved SH adaptive federated filtering method to achieve coordinated control and error correction of multi-source power units, ensuring the coordination and consistency of each power unit when performing tasks, avoiding conflicts and duplication of work, and improving the overall performance of the system.

[0062] 3. The present invention integrates sonar, camera and current meter through the environmental perception module, combined with machine vision technology, which can identify and track oil booms in real time, provide high-precision environmental data, and ensure the stable operation of the system in complex sea conditions.

[0063] 4. The adaptive adjustment unit of the present invention includes a decision-making system based on fuzzy logic or neural network, and a data fusion system based on Kalman filtering or multi-sensor data fusion algorithm. It can dynamically adjust the working status of each power device according to real-time environmental data and task requirements, thereby improving the flexibility and adaptability of the system.

[0064] 5. The dynamic modeling and intelligent decision-making unit of the present invention uses deep learning and reinforcement learning technologies to update model parameters in real time, optimize system performance, and ensure optimal operation in different environments.

[0065] 6. The redundant system unit of the present invention is used to take over the functions when some devices fail, ensuring the reliability and continuity of the system, preventing system shutdown or performance degradation, improving the safety of the system, and significantly improving the efficiency, safety and reliability of offshore oil spill recovery operations. It has broad application prospects and significant economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0067] The following combination Figure 1 The present invention is described in further detail.

[0068] The present invention discloses an intelligent underwater power oil boom traction system based on multi-source collaboration, referring to Figure 1 , including a main controller, an environmental perception unit, a navigation and positioning unit, an adaptive adjustment unit, a dynamic modeling and intelligent decision-making unit, a redundant system unit, an intelligent control unit, a power management unit and multiple underwater power units, among which,

[0069] The underwater power unit is connected to the power management unit, including the propeller, control module, power module and communication module, and each underwater power unit is equipped with a local controller with autonomous decision-making capabilities;

[0070] Specifically, the underwater power device is fixed on the oil containment boom through magnetic connection or mechanical connection, and the controller is a high-efficiency and low-noise electric propeller; the control module is provided with a microprocessor and is equipped with a sensor for obtaining the state of the underwater power device, including speed, position and temperature; the power module uses replaceable high-energy density lithium batteries, the communication module is used for communication with other devices inside the system and the external control center, responsible for sending and receiving data, including state information, control instructions, environmental data, etc., and at the same time, the underwater power devices adopt anti-interference communication protocol for information transmission between each other, to ensure the communication quality in complex underwater environment.

[0071] The main controller is connected with the environment perception unit, responsible for overall task allocation and coordination, to ensure that multiple underwater power devices work cooperatively according to the predetermined strategy;

[0072] Specifically, the main controller includes a cooperative control strategy module, which adopts a consensus algorithm, such as Leader-Follower algorithm, Flocking algorithm, etc., for realizing information exchange and cooperative action among multiple underwater power devices, and through iterative method, the state of each underwater power device gradually tends to be consistent, so as to realize coordinated task execution, and the real-time data provided by the environment perception module is used as the input of the consensus algorithm;

[0073] Its working principle is as follows:

[0074] Information exchange: each underwater power device exchanges state information such as position, speed, direction, etc. with adjacent other devices through the communication module;

[0075] State update: each underwater power device updates its own state according to the received neighbor information using the consensus algorithm, and this process is usually realized through first-order differential equation or difference equation;

[0076] Iterative convergence: through multiple iterations, the states of all underwater power devices gradually converge to a common value or reach a predetermined coordinated state;

[0077] The environment perception unit is connected with the adaptive adjustment unit and the navigation and positioning unit, integrated with sonar, camera and current meter, combined with machine vision technology, for identifying and tracking the oil containment boom;

[0078] The navigation and positioning unit includes acoustic sensors and optical sensors, realizing multi-source sound-light fusion positioning, wherein the acoustic sensors are used for long-distance positioning, and the optical sensors are used for short-distance accurate positioning;

[0079] The adaptive adjustment unit is connected with the dynamic modeling and intelligent decision unit, including a decision system based on fuzzy logic or neural network, and a data fusion system based on Kalman filtering or multi-sensor data fusion algorithm;

[0080] The decision system based on fuzzy logic maps the input values onto fuzzy sets, performs fuzzy operations and reasoning, and finally obtains the output values, and the steps implemented in the application are as follows:

[0081] 1. Input variables: speed, position;

[0082] 2. Membership functions: define fuzzy sets of speed and position, such as "low", "medium", "high";

[0083] 3. Rule base: define If-Then rules, such as "if speed is high and position is off, then slow down";

[0084] 4. Reasoning and defuzzification: perform fuzzy reasoning according to the rule base to obtain the control signal, and defuzzify it to a specific value;

[0085] In the decision system based on neural network, the neural network simulates the working principle of biological neurons, performs complex nonlinear mapping and pattern recognition, has self-adaptive learning ability, and can adjust its parameters according to input data, and the steps implemented in the application are as follows:

[0086] 1. Network structure: input layer (speed, position), hidden layer (several neurons), output layer (control signal);

[0087] 2. Training data: collect speed, position and corresponding control signal;

[0088] 3. Training process: use back propagation algorithm to adjust network weights and biases so that the output is as close as possible to the expected control signal;

[0089] 4. Application: according to new speed and position data, output the corresponding control signal,

[0090] In the data fusion system based on Kalman filtering algorithm, Kalman filtering is a recursive algorithm used to estimate the state of a dynamic system, which combines the dynamic model of the system and the sensor measurement data to provide optimal estimation, and the steps implemented in the application are as follows:

[0091] 1. State prediction: predict the speed and position at the next time according to the dynamic model of the system;

[0092] 2. Covariance prediction: predict the covariance matrix at the next time;

[0093] 3. Measurement update: Combine sensor measurement data to update the estimate of speed and position;

[0094] 4. Iteration: Repeat the above steps to continuously update the estimates of speed and position

[0095] In a data fusion system based on a multi-sensor data fusion algorithm, multi-sensor data fusion improves the perception accuracy and reliability of the system by combining data from multiple sensors. The steps implemented in the present invention are:

[0096] 1. Data preprocessing: denoising and calibrating the data from sonar, camera, and current meter;

[0097] 2. Data fusion: Using Kalman filtering, the data from sonar, camera, and current meter are combined to obtain the optimal speed and position estimates;

[0098] 3. State update: Update the system’s velocity and position estimates based on the fused data;

[0099] 4. Feedback adjustment: Adjust the sensor's operating parameters based on the estimated results of speed and position;

[0100] The dynamic modeling and intelligent decision-making unit is connected to the redundant system unit and the intelligent control unit, including the deep learning model and the reinforcement learning agent, and uses the system recognition model to update the model parameters in real time;

[0101] Specifically, in the dynamic modeling and intelligent decision-making unit, the deep learning model extracts high-level features from the system operation data and provides them to the reinforcement learning agent and the system identification model. The reinforcement learning agent learns the optimal strategy based on the features extracted by the deep learning model, makes real-time decisions and controls the system behavior. The system identification model establishes and updates the dynamic model of the system based on the input and output data of the system and provides it to the deep learning model and the reinforcement learning agent. The adaptive adjustment unit combines deep learning, reinforcement learning and system identification technologies to adjust the working state of the system in real time and improve the flexibility and adaptability of the system, including:

[0102] 1. Deep learning model training:

[0103] Collect sensor data and control signals during system operation, design a convolutional neural network (CNN) to extract spatial features from the data, train the CNN using the collected data, adjust the network weights using a backpropagation algorithm, deploy the trained CNN into the system, and predict the system status in real time.

[0104] 2. Reinforcement Learning Agent Strategy Optimization:

[0105] Define the state space (such as position, velocity), action space (such as acceleration, deceleration) and reward function (such as minimizing energy consumption) of the system, design a deep Q network (DQN) to learn the optimal policy, in the simulation environment, the agent adjusts the policy through trial and error to maximize the cumulative reward, uses gradient descent algorithm to optimize the parameters of DQN, the trained DQN strategy is deployed in the system to make real-time decisions and control system behavior;

[0106] 3. System identification model update:

[0107] Collect the input and output data of the system, select a state space model to describe the dynamic behavior of the system, use Kalman filtering algorithm to estimate the model parameters, use independent data set to verify the model performance, update the model parameters in real time during the system operation to adapt to the dynamic changes of the system;

[0108] Redundant system units, which are divided into hardware redundancy and software redundancy, are used to take over the function when part of the device fails, and the backup component takes over the function of the failed component to avoid system downtime or performance degradation, ensuring the continuity and reliability of the system;

[0109] Specifically, two controllers are configured in the system, one as the main controller, and the backup controller is set in the redundant system unit, the backup controller is connected with the main controller through high-speed communication link, the main controller is responsible for normal operation, and the backup controller is in standby state, the main controller and the backup controller are synchronized in real time through high-speed communication link to ensure that the backup controller can take over at any time; The main and backup switching algorithm is set in the control system, when the main controller fails, the backup controller automatically takes over its function, the system failure is detected in real time through sensors and monitoring system, and the failed components are isolated to prevent fault propagation;

[0110] The system is configured with a fault detection module to monitor the status of the main controller in real time, and when the main controller fails, it is immediately switched to the backup controller, and multiple sensors are configured at key positions to ensure that even if one sensor fails, the system can still obtain accurate data, multiple power modules are configured to ensure that even if one power module fails, the system can still be powered normally, and multiple communication paths are configured to ensure that even if one path fails, data can still be transmitted through other paths;

[0111] The intelligent control unit is connected with the power management unit, and the improved S-H adaptive federated filtering method is used to monitor the status of each underwater power device in real time, and adjust the working state according to the actual situation, realizing the collaborative control and error correction of multi-source underwater power devices;

[0112] Specifically, in the intelligent control unit, the S-H adaptive federated filtering method is an algorithm for multi-source data fusion and error correction, the implementation steps are:

[0113] 1. Model the underwater multi-source integrated navigation system to obtain the error model of each navigation sensor;

[0114] 2. Propose an improved S-H adaptive filtering method based on the error model;

[0115] 3. In the federated filter, use the improved S-H adaptive filtering method for time update and measurement update;

[0116] 4. Through the federated filter, fuse the data of multiple sensors to obtain the optimal state estimation;

[0117] The implementation steps of the intelligent control unit are:

[0118] 1. Each underwater power device is equipped with sensors to collect real-time data such as position, speed, and power, and transmits the sensor data to the intelligent control unit through the communication module. The intelligent control unit monitors the status of each underwater power device in real time, analyzes the data, and identifies abnormal conditions;

[0119] 2. The intelligent control unit sends control instructions to each underwater power device to adjust the operating state of the thruster. Each underwater power device adjusts its own state according to the received control instructions and feeds back the new state information to the intelligent control unit to form a closed-loop control;

[0120] The specific process is:

[0121] 1. System modeling: Model the underwater multi-source integrated navigation system to obtain the error model of each navigation sensor. The error model of the position sensor can be expressed as:

[0122] x k+1 =Ax k +Bu k +w k

[0123] where x k is the state vector, u k is the control input, and w k is the process noise;

[0124] 2. Error correction: Use the improved S-H adaptive filtering method to correct the sensor data errors. Use Kalman filtering for state estimation and error correction:

[0125]

[0126] where k k is the Kalman gain, z k is the measurement value, and H is the measurement matrix;

[0127] 3. Cooperative control: Consensus algorithm is used to achieve cooperative control of multiple underwater power devices, and Leader-Follower algorithm is used:

[0128]

[0129] where u i is the control input of the i-th device, N i is the neighbor set of the i-th device, a ij is the adjacency matrix element;

[0130] 4. Real-time monitoring and state adjustment: The intelligent control unit monitors the state of each underwater power device in real time, analyzes the data, identifies abnormal conditions, and sends control instructions to each underwater power device according to the calculation results, adjusts the running state of the propeller, and each underwater power device adjusts its state according to the received control instructions, and feeds back the new state information to the intelligent control unit, forming a closed loop control;

[0131] Power management unit, real-time monitoring of the power state of each underwater power device.

[0132] In further embodiments, the system of the present application further comprises:

[0133] Data storage module for storing system operation data and sensor data;

[0134] Fault diagnosis module for detecting system faults and providing recommended solutions;

[0135] Simulation training module for simulating different operating environments and fault conditions to train the system response:

[0136] Remote control module allows operators to remotely monitor and control the operation of the system;

[0137] Modular design and standardized interface for quick assembly, maintenance and upgrade of the system;

[0138] Software configuration system based on XML or JSON, using international standard-based communication and control interface such as ROS (Robot Operating System) framework.

[0139] Method for using intelligent underwater power boom drag system based on multi-source cooperation, comprising the following steps:

[0140] S1, system initialization, the main controller assigns specific tasks to each underwater power device according to the preset task plan, including initial position, path planning and working mode;

[0141] Specifically, at system startup, each underwater power device establishes a connection with the main controller through the wireless communication module, ensuring the stability and real-time nature of data transmission. The main controller allocates specific tasks to each power device based on the pre-set task plan, including initial position, path planning, and working mode. The environmental perception module is started, and real-time surrounding environmental data (such as water flow speed, direction, temperature, etc.) are collected and transmitted to the navigation and positioning unit and the adaptive adjustment unit.

[0142] S2, deployment phase, during the inflation and deployment of the oil containment boom, the operator or automatic equipment installs underwater power devices on the oil containment boom at predetermined intervals to ensure uniform distribution of the power devices. Each power device exchanges information with the main controller and other power devices through the communication module, ensuring synchronization of the status and position of each device. The environmental perception module monitors the position and status of the oil containment boom in real time, providing high-precision data to support decision-making by the navigation and positioning unit and the adaptive adjustment unit.

[0143] S3, dragging phase, each underwater power device adjusts the thrust and direction of the propeller according to the instructions of the main controller, cooperatively providing sufficient traction to ensure that the oil containment boom operates within the predetermined area. Meanwhile, the navigation and positioning unit provides high-precision positioning information by combining data from acoustic sensors and optical sensors, ensuring accurate deployment and dragging of the oil containment boom. The adaptive adjustment unit dynamically adjusts the working state of each power device based on real-time environmental data (such as water flow speed, direction, etc.) and task requirements, ensuring the flexibility and adaptability of the system. The intelligent control unit uses the improved S-H adaptive federated filtering method to correct errors in the status of each power device, ensuring the coordination and consistency of each device.

[0144] S4, charging and battery replacement, the power management unit monitors the power state and state of charge of each underwater power device in real time through voltage, current, and temperature sensors. When the power is below the set threshold, the power management unit starts the charging process, controlling the charging device to perform constant current charging, constant voltage charging, or pulse charging, ensuring that the battery is charged within a safe range. When the state of charge of the battery is below the set threshold, the power management unit prompts the need for battery replacement and coordinates the battery replacement process, including battery removal, installation, and initialization.

[0145] S5, fault handling and redundancy management, the redundancy system unit monitors the status of each underwater power device in real time. When a device failure is detected, the redundancy system immediately takes over its functions.

[0146] S51, the intelligent control unit analyzes sensor data on the underwater power device, such as voltage, current, temperature, position, speed, etc., and identifies abnormal conditions, such as low voltage, high temperature, or position deviation. When an abnormal condition is detected, the system issues a fault alarm and records the fault information.

[0147] S52, real-time synchronization of data between the main controller and the backup controller, ensuring that the backup controller can take over at any time, and when the main controller fails, the backup controller automatically takes over its functions, ensuring continuous operation of the system, implementing redundant control algorithms and data synchronization, ensuring that the backup controller can seamlessly take over when the main controller fails, redundant control algorithms include main-backup switching algorithms, fault detection and isolation algorithms, etc.;

[0148] S53, real-time detection of system failures through sensors and monitoring systems, and isolation of failed components to prevent fault propagation;

[0149] S54, the redundant system unit takes over the functions of the failed components, ensuring the reliability and continuity of the system, and records fault information and prompts the operator to repair the fault;

[0150] S55, after repair is completed, the system is reinitialized and resumes normal operation, and the redundant system unit synchronizes data during the fault period to the main controller, ensuring data consistency and integrity;

[0151] S6, dynamic modeling and intelligent decision-making, the dynamic modeling and intelligent decision-making unit uses deep learning and reinforcement learning techniques to update model parameters in real time and optimize system performance;

[0152] S61, collect sensor data during system operation, including position, speed, power, etc., and collect control signals during system operation, including thruster control instructions, etc.;

[0153] S62, clean, normalize and segment the collected data to prepare training and test data sets;

[0154] S63, design the structure of a deep neural network, common network structures include convolutional neural networks (CNN) and recurrent neural networks (RNN), including input layer, multiple hidden layers and output layer, use training data sets, adjust network weights and biases through backpropagation algorithm, so that the model can accurately predict the state and behavior of the system;

[0155] S64, use test data sets to verify the performance of the model, adjust hyperparameters to optimize the model, and deploy the trained model to the system to predict and adjust the system state in real time;

[0156] S65, define the state space, action space and reward function of the system, build a simulation environment, design a reinforcement learning agent, including policy network and value network, common algorithms include deep Q network (DQN), policy gradient (PG) and proximal policy optimization (PPO);

[0157] S66, the intelligent agent explores and learns in the simulation environment, continuously adjusts the strategy through trial and error to maximize the cumulative reward;

[0158] S67, using gradient descent and other optimization algorithms, adjust the parameters of the strategy network to improve the performance of the strategy, deploy the trained strategy to the system, make real-time decisions and control system behavior;

[0159] S68, collect the input and output data of the system, including sensor data and control signals. Select appropriate model structure, such as linear model, nonlinear model or neural network model, use least squares method, Kalman filter and other algorithms to estimate model parameters, so that the model can accurately describe the dynamic behavior of the system, use independent data set to verify the performance of the model, adjust the model structure and parameters to optimize the model, update the model parameters in real time during the system operation to adapt to the dynamic changes of the system;

[0160] S7, system recycling, after the task is completed, the main controller sends the recycling instruction, each underwater power device stops working and returns to the designated position;

[0161] S71, the main controller sends the recycling instruction to each underwater power device according to the task completion condition, each underwater power device stops the current work after receiving the recycling instruction, and prepares to return to the designated position;

[0162] S72, the navigation and positioning unit plans the optimal recycling path according to the current environment and the position of each underwater power device, and the self-adaptive adjustment unit dynamically adjusts the recycling path of each underwater power device according to the real-time environmental data and task requirements;

[0163] S73, each underwater power device coordinates the return through the consistency algorithm to avoid conflict and repeated work, and the intelligent control unit monitors the state of each underwater power device in real time to ensure the safety and efficiency of the recycling process;

[0164] S74, after each underwater power device returns to the designated position, personnel or automatic equipment recycle and maintain the underwater power device, the power management unit checks the power state of each underwater power device, and performs necessary charging or battery replacement, the system records the data in the recycling process, and analyzes and optimizes it for the next use.

[0165] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, which should be covered by the claims of the present application.

Claims

1. The intelligent underwater power boom traction system based on multi-source collaboration is characterized by: It includes a main controller, an environmental perception unit, a navigation and positioning unit, an adaptive adjustment unit, a dynamic modeling and intelligent decision-making unit, a redundant system unit, an intelligent control unit, a power management unit and multiple underwater power units, among which, The underwater power unit is connected to the power management unit, including the propeller, control module, power module and communication module, and each underwater power unit is equipped with a local controller with autonomous decision-making capabilities; The main controller is connected to the environmental perception unit and is responsible for overall task allocation and coordination, ensuring that multiple underwater power units operate in coordination according to predetermined strategies; The environmental perception unit is connected to the adaptive adjustment unit and the navigation and positioning unit, and is integrated with sonar, cameras and current meters for identifying and tracking oil booms; The navigation and positioning unit includes an acoustic sensor and an optical sensor to achieve multi-source acoustic and optical fusion positioning. The acoustic sensor is used for long-distance positioning, and the optical sensor is used for short-distance precise positioning. The adaptive adjustment unit is connected with the dynamic modeling and intelligent decision-making unit, including a decision-making system based on fuzzy logic or neural network, and a data fusion system based on Kalman filtering or multi-sensor data fusion algorithm; The dynamic modeling and intelligent decision-making unit is connected to the redundant system unit and the intelligent control unit, including the deep learning model and the reinforcement learning agent, and uses the system recognition model to update the model parameters in real time; Redundant system units, used to take over functions when some devices fail; The intelligent control unit is connected to the power management unit to monitor the status of each underwater power unit in real time and adjust the working status according to the actual situation, realizing the coordinated control and error correction of multi-source underwater power units; The power management unit monitors the power status of each underwater power device in real time.

2. The intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 1 is characterized in that: The main controller includes a collaborative control strategy module, which adopts a consistency algorithm to realize information exchange and collaborative action among multiple underwater power devices. The real-time data provided by the environmental perception module serves as the input of the consistency algorithm.

3. The intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 1 is characterized in that: A backup controller is provided in the redundant system unit, and the backup controller is connected to the main controller via a high-speed communication link.

4. The intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 1 is characterized in that: The control module in the underwater power device is equipped with a microprocessor and a sensor for obtaining the status of the underwater power device. The power module uses a replaceable lithium battery, and the communication module is used to communicate with other devices within the system and an external control center.

5. The intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 1 is characterized in that: Also includes: A data storage module for storing system operation data and sensor data; Fault diagnosis module, which detects system faults and provides suggested solutions; Simulation training module, used to simulate different operating environments and fault conditions to train system responses: Remote control module, which allows operators to remotely monitor and control the operation of the system; Modular design and standardized interfaces enable rapid assembly, maintenance, and upgrade of the system; The software configures the system, using communication and control interfaces based on international standards.

6. The method for using the intelligent underwater power oil boom traction system based on multi-source collaboration is characterized in that: The following steps are involved: S1, system initialization, the main controller assigns specific tasks to each underwater power unit according to the preset mission plan, including initial position, path planning and working mode; S2. During the boom deployment process, operators or automated equipment install underwater propulsion devices on the boom at predetermined intervals; S3. Each underwater power unit adjusts the thrust and direction of the propeller according to the instructions of the main controller, and collaboratively provides sufficient traction to enable the oil boom to operate within the predetermined area; S4. The power management unit monitors the power status and charge status of each underwater power device in real time through voltage, current and temperature sensors; S5, the redundant system unit monitors the status of each underwater power device in real time. When a device failure is detected, the redundant system is immediately activated to take over its function; S6, dynamic modeling and intelligent decision-making unit uses deep learning and reinforcement learning technology to update model parameters in real time and optimize system performance; S7. After the mission is completed, the main controller issues a recovery command, and each underwater power unit stops working and returns to the designated location.

7. The method for using the intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 6 is characterized in that: In step S4, when the power level is lower than the set threshold, the power management unit starts the charging process and controls the charging equipment to perform constant current charging, constant voltage charging or pulse charging to ensure that the battery is charged within a safe range. When the battery's state of charge is lower than the set threshold, the power management unit prompts that the battery needs to be replaced and coordinates the battery replacement process.

8. The method for using the intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 6 is characterized in that: Step S5 includes: S51, the intelligent control unit analyzes sensor data on the underwater power device and identifies abnormal conditions. When an abnormal condition is detected, the system issues a fault alarm and records fault information; S52, real-time data synchronization between the main controller and the backup controller to ensure that the backup controller can take over at any time. When the main controller fails, the backup controller automatically takes over its functions; S53, using sensors and monitoring systems to detect system failures in real time and isolate failed components to prevent the spread of failures; S54, the redundant system unit takes over the function of the failed component to ensure the reliability and continuity of the system. The system records the fault information and prompts the operator to repair the fault; S55. After the repair is completed, the system is reinitialized and resumes normal operation. The redundant system unit synchronizes the data during the fault period to the main controller.

9. The method for using the intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 6 is characterized in that: Step S6 includes: S61, collecting sensor data during system operation and collecting control signals during system operation; S62. Clean, normalize and segment the collected data to prepare training and test datasets; S63. Design the structure of a deep neural network, including an input layer, multiple hidden layers, and an output layer. Use a training dataset and backpropagation to adjust the network weights and biases so that the model can accurately predict the state and behavior of the system. S64. Use the test dataset to verify the performance of the model, adjust the hyperparameters to optimize the model, deploy the trained model to the system, and predict and adjust the system status in real time; S65. Define the system's state space, action space, and reward function, build a simulation environment, and design a reinforcement learning agent, including a policy network and a value network. S66, enables the agent to explore and learn in the simulation environment, and continuously adjusts the strategy through trial and error to maximize the cumulative reward; S67. Use optimization algorithms such as gradient descent to adjust the parameters of the policy network to improve the performance of the policy. Deploy the trained policy to the system to make real-time decisions and control the system behavior. S68. Collect the input and output data of the system, including sensor data and control signals, select the appropriate model structure, estimate the model parameters so that the model can accurately describe the dynamic behavior of the system, verify the performance of the model using independent data sets, adjust the model structure and parameters to optimize the model, and update the model parameters in real time during system operation to adapt to the dynamic changes of the system.

10. The method for using the intelligent underwater power oil boom traction system based on multi-source collaboration according to claim 6, characterized in that: Step S7 includes: S71. The main controller sends a recovery command to each underwater power unit according to the task completion status. After receiving the recovery command, each underwater power unit stops the current operation and prepares to return to the designated location. S72. The navigation and positioning unit plans the optimal recovery path based on the current environment and the position of each underwater power unit. The adaptive adjustment unit dynamically adjusts the recovery path of each underwater power unit based on real-time environmental data and mission requirements. S73. Each underwater power unit coordinates its return through a consistency algorithm to avoid conflicts and duplication of work. The intelligent control unit monitors the status of each underwater power unit in real time. S74. After each underwater power unit returns to the designated location, personnel or automatic equipment will recover and maintain the underwater power unit. The power management unit will check the power status of each underwater power unit and perform necessary charging or battery replacement. The system will record the data during the recovery process, analyze and optimize it, and prepare for the next use.

Images