Intelligent electronic anchoring control method and system based on ship

Abstract

The invention provides an intelligent electronic anchoring control method and system based on a ship, and relates to the technical field of unmanned ship autonomous control, and the method comprises the following steps: S1, activating an electronic anchoring function, obtaining a matched initial anchoring parameter and an initial anchoring target area according to a current operation mode, wherein the initial anchoring parameters comprise an anchoring radius R, an allowable deviation threshold value epsilon and MPC initial parameters; determining the ship position at the activation moment as the center of a target area according to the initial anchoring target area; s2, ship state data, environment interference data and obstacle data are collected according to the initial anchoring target area, redundant fusion processing is carried out on the collected multi-source data, and fusion data are output. Compared with an existing method, a position complementing scheme can be automatically calculated through preset thrust distribution logic when a part of the propeller breaks down; through dynamic interference evaluation and adaptive adjustment of MPC parameters, precision and energy consumption can be dynamically balanced according to environmental interference, and the load of the propeller is reduced when interference is low.

Description

Technical Field

[0001] This invention relates to the field of autonomous control technology for unmanned surface vessels, and in particular to a smart electronic anchoring control method and system based on vessels. Background Technology

[0002] With the rapid development of water recreation scenarios, the frequency of use of small boats such as fishing boats and leisure yachts has increased significantly. When these boats are operating or moored, the demand for "electronic anchoring function" is becoming increasingly urgent. Electronic anchoring needs to maintain the boat's fixed position in a designated water area by dynamically adjusting the thruster without the need for physical anchoring. It needs to ensure positional accuracy to meet usage requirements, while also taking into account the limited energy reserves of small boats to avoid shortening the range. At the same time, it also needs to cope with the complex dynamic environment common in nearshore waters, such as sudden gusts of wind of 3-5 m / s and sudden changes in water flow.

[0003] In existing technologies, CN111221336B proposes a ship-side assisted berthing and unberthing system, which achieves automated control of the berthing process through decision-making modules and sensing devices mounted on the ship itself; CN113009909A develops a ship-shore cooperative single-ship autonomous berthing and mooring system, which uses a ship-shore cooperative mechanism to complete the autonomous berthing and mooring of the ship. These systems provide preliminary support for the automated berthing of small vessels, but they have not effectively solved the problem of long-term stable operation under electronic anchoring conditions. For example, the CN111221336B system only focuses on position control during berthing and lacks an adaptive adjustment mechanism for accuracy and energy consumption in the anchoring state. This causes the propeller of small boats to run at full load continuously during long-term anchoring, resulting in a significant reduction in range. The CN113009909A system relies on fixed parameter control. When faced with sudden gusts of wind or changes in water flow such as 3-5 m / s, the anchoring deviation exceeds 1.5 m. Moreover, the MPC control uses a fixed weight matrix and cannot dynamically adjust the accuracy priority according to environmental interference. Both systems have insufficient accuracy compensation under high interference, leading to anchoring deviation. Maintaining high accuracy output under low interference results in energy waste. They cannot balance the accuracy and energy consumption requirements of electronic anchoring for small boats. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, the technical problem to be solved by this invention is to provide a ship-based intelligent electronic anchoring control method and system that can achieve a dynamic adaptive balance between anchoring accuracy and energy consumption in complex dynamic environments, avoiding energy waste caused by excessive pursuit of accuracy under low interference conditions and excessive anchoring deviation caused by fixed parameters under high interference conditions.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: The present invention provides a ship-based intelligent electronic anchoring control method, comprising the following steps: S1. Activate the electronic anchoring function, obtain the matching initial anchoring parameters and initial anchoring target area according to the current operating mode, where the initial anchoring parameters include the anchoring radius R, the allowable deviation threshold ε, and the initial MPC parameters; determine the target area centered on the ship's position at the time of activation based on the initial anchoring target area. S2. Collect vessel status data, environmental interference data, and obstacle data based on the initial anchoring target area. Perform redundant fusion processing on the collected multi-source data and output fused data. S3. Calculate the environmental interference level based on the environmental interference data in the fused data. According to the level of environmental interference In step S1, dynamically adjust the initial MPC parameters and output the adjusted MPC parameters; in step S4, based on the ship state data in the fused data and the adjusted MPC parameters, construct an optimization objective function with energy consumption constraints and a ship dynamics model, and solve for the propeller control variables. S5. Based on the thruster control quantity The thrust is distributed to each electric thruster according to preset rules. When a thruster malfunction is detected, the dynamic adjustment of the distribution ratio is triggered, and a thrust command is output; S6, Collect the actual position of the ship after the thrust command is executed. Calculate the actual position of the ship Center of the target area deviation According to the deviation The logic triggers the return to step S2 or step S3; S7, real-time monitoring of the running status of steps S2-S6, triggering an alarm and executing a backup strategy when a fault is detected, while recording and uploading fault data to form a closed-loop control.

[0006] In the preferred embodiment, step S2 involves deploying a positioning and environmental perception module based on the initial anchoring target area determined in step S1. This module includes a positioning sensor, an inertial measurement unit, a wind direction sensor, a water flow sensor, and an obstacle detection sensor. The specific steps are as follows: S21. The positioning and environmental perception module collects ship status data, environmental interference data, and obstacle data at a preset frequency; the ship status data includes position... Heading angle yaw angle Pitch angle Longitudinal velocity and lateral speed Environmental interference data includes wind speed (W) and wind direction. Water flow velocity C, water flow direction Obstacle data refers to the distance information of obstacles within 1m to 30m around the ship. S22. The data acquired in step S21 is fused using an improved Kalman filter algorithm. The state vector of the improved Kalman filter algorithm is defined as follows: ;in, For the ship's position, These are the longitudinal and lateral velocities, respectively. These are the roll and pitch angles, respectively; the observation matrix of the improved Kalman filter algorithm is... Observation matrix The state vector A linear mapping is used to obtain the predicted observation vector at the current moment; the status of the positioning sensor is acquired; if the data update interval of the positioning sensor is ≤0.1s, it is considered normal; if the data interruption of the positioning sensor is >5s, it is considered faulty; the observation matrix is ​​dynamically adjusted according to the status of the positioning sensor. : When the positioning sensor determines that it is normal, then the observation matrix... Set as: , observation ; When the positioning sensor determines that it has failed, the observation matrix... Set as: Velocity integral completion via inertial measurement unit ; S23. The data fused in step S22 is output to steps S3 and S6 respectively at a preset frequency.

[0007] In the preferred embodiment, the initial parameters for MPC in step S3 include the initial prediction step size. Initial position weight matrix Initial control weight matrix and initial energy consumption weight The specific steps of step S3 are as follows: S31. Based on the wind speed W and water flow velocity C in the environmental interference data output in step S2, introduce the historical interference attenuation factor. Calculate the current environmental interference level : ;in, The interference level at the previous moment. , ; S32. Environmental interference level calculated based on step S31 Adjust the position weight matrix Q and energy consumption weight in the initial parameters of MPC. :like If so, it is determined to be low interference: , ;in, This is the initial position weight matrix in step S1. As the initial energy consumption weight; if If so, it is determined to be a medium-level interference: , ;like If so, it is determined to be high interference: , ; S33, Adjust Q after step S32. and the initial prediction step size in step S1 and the initial control weight matrix Perform the integration operation to obtain the adjusted MPC parameters and output them to step S4.

[0008] In the preferred embodiment, in step S4, based on the vessel state data and environmental disturbance data output in step S2, and combined with the inherent parameters of the small vessel, the following three-dimensional dynamic model is constructed: Longitudinal position prediction equation: ; in, t represents the control period; t represents the current control time; k represents the prediction step index of the MPC algorithm. Let be the longitudinal position coordinates of the ship at time t+k; Let be the longitudinal velocity of the ship at time t+k; Let be the total longitudinal thrust of the ship at time t+k; This refers to the longitudinal force exerted by the wind on the ship. ρ represents the longitudinal force exerted by the water flow on the ship; m represents the mass of the ship. The predicted longitudinal position of the ship at time t+k+1; Lateral position prediction equation: ; in, Let be the lateral position coordinates of the ship at time t+k; Let be the lateral velocity of the ship at time t+k; Let be the total lateral thrust of the ship at time t+k; The lateral force exerted by the wind on the ship; The lateral force exerted by the water flow on the ship; The predicted lateral position of the ship at time t+k+1; and ; ; Heading angle prediction equation: ; in, Let be the heading angle of the ship at time t+k; Let be the angular velocity of the ship at time t+k; Let t+k be the bow turning moment of the ship; The turning moment of the ship due to wind; I is the turning torque of the water flow on the ship; I is the moment of inertia of the ship about its vertical axis. The predicted heading angle of the ship at time t+k+1; and ; ; Bow torque equation: ; in, For the turning torque, Let A be the density of seawater and A be the surface area of ​​the waterline. , These are the longitudinal and lateral drag coefficients, respectively. , These are the thrusters at the stern and bow, respectively. , B is the distance from the propeller to the center of the ship, and B is the width of the ship. , These are the thrusters on the right and left sides of the ship, respectively.

[0009] In the preferred embodiment, step S4 further includes the following specific steps: S41. Construct the optimization objective function based on the adjusted MPC parameters: ; in, For time t Predicted location at time The initial anchoring target area center in step S1, for The generalized control input of the time-of-flight thruster is defined by N, where N is the MPC prediction step size, and Q and R are the position weight matrix and control input weight matrix, respectively. Energy consumption weighting coefficient This is an energy consumption constraint term; S42. Constrain the amplitude and anchoring range of the control quantity. Solve the optimization objective function of step S41 using a quadratic programming algorithm to obtain the thruster control quantity at the current moment. And output it to step S5; The control amplitude constraint is as follows: ,in For the minimum value of the thruster control quantity, Set to the maximum value of the thruster control quantity; The anchoring range constraint is: , where R is the anchoring radius determined in step S1.

[0010] In the preferred embodiment, step S5 is as follows: S51. Output from step S4 The electric propulsion units, including the bow propulsion unit, stern propulsion unit, port propulsion unit, and starboard propulsion unit, are allocated according to the following rules: Bow thrusters: The thrust direction and the heading angle output in step S2 Consistent; Stern thruster: Thrust direction and Consistent; Left thruster: Thrust direction and Vertically to the left; Right-side thruster: Thrust direction and Vertically to the right; S52. Obtain the load and response status of each thruster; if a single thruster fails, trigger the adjustment of the lateral thrust distribution ratio; if both thrusters fail, retain the longitudinal thrust and trigger the emergency handling in step S7. S53. Convert the thrust of each propeller determined in step S51 or S52 into a corresponding speed command, output it to the propeller for execution, and after execution, adjust the ship's position and feed back the status to step S2.

[0011] In the preferred embodiment, step S6 is as follows: S61. Collect the actual position of the ship after step S5 is executed according to the preset frequency. ,calculate With respect to the center of the initial anchored target area in step S1 Euclidean deviation ; S62, if Allowable deviation threshold If the current thruster control value is maintained, return to step S2 to continue data collection; if Allowable deviation threshold Then Feedback is sent to step S3, and steps S3-S5 of MPC parameter adjustment and thrust control are re-executed.

[0012] In the preferred embodiment, step S7 is as follows: S71. Monitor the sensor data in step S2, the time taken to solve the control quantity in step S4, and the thruster fault status in step S5 in real time. If the sensor data interruption exceeds 5 seconds, the time taken to solve the control quantity exceeds 1 second, or the thruster fails, it is determined to be a fault. S72. Trigger an audible and visual alarm when a fault occurs. If a sensor fault is confirmed, enable the redundant fusion data in step S2. If a thruster malfunction is determined, the adjustment strategy in step S52 is executed; If a serious malfunction is identified, all thruster thrust is cut off and mechanical braking is activated. S73. Record key data at the time of the fault. Key data includes the fused data from step S2 and the thruster control quantity from step S4. Deviation in step S6 Key data is backed up locally and then uploaded to the server.

[0013] In a preferred embodiment, a system for implementing the ship-based intelligent electronic anchoring control method described in any of the above embodiments is also provided, comprising: The positioning and environmental perception module is used to perform multi-source data acquisition in step S2 of claim 1, including a positioning sensor, an inertial measurement unit, a wind direction sensor, a water flow sensor and an obstacle detection unit, and outputs raw acquisition data; The intelligent control module is used to execute the parameter calling in step S1 of claim 1, the redundancy fusion in step S2, the parameter adjustment in step S3, and the control quantity solution in step S4, including an MPC control law solution unit, a redundancy filtering unit, an adaptive weight adjustment unit, and a parameter storage unit. A power adjustment and execution mechanism, used to perform the thrust distribution and command execution in step S5 of claim 1, includes at least a plurality of electric thrusters, a thrust distribution unit and a thruster drive unit; An emergency handling module is used to perform the fault handling and log management of step S7 of claim 1, and includes at least a fault detection unit, an emergency stop response unit, a log recording unit, and an alarm unit; The data interaction module is used to realize the function activation in step S1, the log upload and anchor status synchronization in step S7, and includes at least an Ethernet unit and a wireless communication unit. The positioning and environmental perception module, intelligent control module, power regulation and actuator, emergency response module, and data interaction module are sequentially connected by signals to form a closed-loop system of perception, calculation, control, execution, and monitoring.

[0014] In the preferred embodiment, the positioning sensor in the positioning and environmental perception module is a GNSS receiver; the inertial measurement unit is a six-axis IMU; and the obstacle detection unit is a millimeter-wave radar. The hardware carrier of the intelligent control module is an embedded industrial computer, which pre-stores the initial parameter mapping table of step S1 and the ship dynamics model parameters of step S4 in the above scheme, and integrates the OSQP quadratic programming solver in the software. The emergency response unit in the emergency handling module integrates an electromagnetic brake to bring the vessel to a stop within 2 seconds.

[0015] This invention provides a ship-based intelligent electronic anchoring control method and system. Through the coordination of the aforementioned structures, it offers the following advantages compared to existing methods: First, the combination of accurate identification of ship dynamics model parameters and improved Kalman filtering can automatically calculate the replacement scheme through preset thrust distribution logic when the propeller fails. Stable anchoring can be quickly restored without manual intervention, which improves the accuracy of ship state estimation. At the same time, it can also avoid the propeller from running at full load continuously in the anchored state, thus extending the endurance. Secondly, through dynamic interference assessment and adaptive adjustment of MPC parameters, it not only balances accuracy and energy consumption, but also addresses the problem of limited range. It can dynamically balance accuracy and energy consumption according to environmental interference. In low interference, it reduces the thruster load, i.e., reduces unnecessary thrust output. Compared with the full-load operation of existing technologies, it can extend the anchoring time after a single charge. It not only prevents the anchoring deviation from exceeding the standard due to fixed parameters under high interference, but also reduces excessive energy consumption in low interference by dynamically adjusting the energy consumption weight, effectively solving the "range anxiety" problem in long-term anchoring of small ships. Third, the optimized data fusion algorithm and thrust allocation strategy enhance the system's robustness in various environments, effectively balance anchoring accuracy and energy consumption requirements, and solve the problems of insufficient accuracy compensation under high interference and energy waste under low interference in traditional systems. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the invention; Figure 2 This is a visualization of the thrust distribution of the propeller in this invention; Figure 3 This is a diagram illustrating a malfunction scenario of the right-side thruster in this invention. Figure 4 This is a schematic diagram of the system operation interface of the present invention.

[0017] The diagram shows: Positioning and Environmental Perception Module 1, Positioning Sensor 11, Inertial Measurement Unit 12, Wind Direction Sensor 13, Water Flow Sensor 14, Millimeter Wave Radar 15, Intelligent Control Module 2, MPC Control Law Solving Unit 21, Redundant Filtering Unit 22, Adaptive Weight Adjustment Unit 23, Parameter Storage Unit 24, Power Adjustment and Actuation Mechanism 3, Electric Thruster 31, Bow Thruster 311, Stern Thruster 312, Portal Thruster 313, Starboard Thruster 314, Thrust Distribution Unit 32, Thruster Drive Unit 33, Emergency Handling Module 4, Fault Detection Unit 41, Emergency Stop Response Unit 42, Log Recording Unit 43, Audible and Visual Alarm 44, Data Interaction Module 5, Ethernet Module 51, 4G Module 52, Bubble Pig Software 6, Bubble Pig Server 7. Detailed Implementation

[0018] To better understand the purpose, system architecture, and functional implementation of this embodiment, the embodiments and features described herein can be combined with each other without conflict. The exemplary embodiments disclosed herein will be described below with reference to the accompanying drawings, including specific technical details disclosed to aid understanding; however, these details should be considered exemplary rather than restrictive. Therefore, those skilled in the art should understand that various improvements and adjustments can be made to the embodiments described herein without departing from the scope and core ideas of the invention. Similarly, for clarity, detailed descriptions of well-known technologies, functions, and structures are omitted in the following description.

[0019] Example 1 This embodiment details the training process of the MPC control model based on real ship test data, specifically including the calibration of core ship dynamics parameters and the optimization of the noise covariance matrix of the improved Kalman filter. The data acquisition, model building, and parameter optimization steps in the field of ship control in this embodiment are all based on actual ship test records.

[0020] Model Predictive Control (MPC) is an advanced control algorithm based on predictive models, rolling optimization, and feedback correction. In this scheme, the pre-stored ship dynamics model (such as the parameters calibrated in Example 1) is used to predict the ship's position change trend under disturbances such as wind and current over a period of time. Then, the optimal propeller control quantity at the current moment is solved by optimizing the objective function. Finally, the prediction deviation is corrected by combining the real-time feedback ship position data to achieve dynamic adjustment.

[0021] Compared to traditional fixed-parameter control (such as PID), MPC can adjust the weights adaptively, prioritizing anchoring accuracy under high disturbances and reducing energy consumption under low disturbances. This effectively compensates for the shortcomings of the fixed parameters in the existing technology (publication number CN113009909A) which cannot adapt to dynamic scenarios.

[0022] like Figure 1 As shown, the intelligent electronic anchoring control method based on a vessel provided in this embodiment first selects an experimental vessel, a small sightseeing boat with a length of 8m and a width of 2.5m, a full load mass of m=1000kg, a design draft of 0.8m, and is equipped with 4 Torqeedo Cruise10.0R electric thrusters (each with a rated thrust of 500N and a thrust measurement accuracy of...). .

[0023] Specifically, the sensor configuration is shown in Table 1 below: Table 1

[0024] In this embodiment, the test environment is as follows: Water area: Inland sea area, i.e., water depth 3-5m, wave height To avoid interference from extreme environments; Interference scenario coverage: Low interference, i.e., wind speed And water flow speed Interference in the middle, i.e. and High interference, i.e. or ; Experiment duration: 80 hours in total, divided into 20 independent groups, each group lasting 4 hours; Effective data volume: Nearly 800,000 sets of effective data were collected, with 40,000 points in each experiment.

[0025] In this embodiment, the data from the IMU (100Hz), wind direction and speed, and water flow sensor (5Hz) are synchronized to 10Hz (GNSS sampling rate) using linear interpolation based on the GNSS timestamp. Post-synchronization verification: Calculate the deviation between the GNSS position and the IMU integral position. If the deviation... If the data for that period is removed, the actual removal rate is approximately 3%.

[0026] In this embodiment, scene segmentation and labeling According to wind speed and water flow velocity The data is divided into three scenarios: Low interference: 40% of the data, i.e. and Interference: 45% of the data, i.e. and High interference: 15% of the data, i.e. or .

[0027] In this embodiment, noise suppression and outlier handling are specifically as follows: A moving average filter with a length of 5 was applied to the wind speed and water flow velocity data to eliminate instantaneous pulse interference; If the wind speed suddenly changes at a certain moment (For example, short-term gusts of wind), then replace the point with the average of the three times before and after; The final amount of valid data retained is 776,000 sets, which can cover all interference scenarios.

[0028] In this embodiment, the calibration of the core parameters of ship dynamics is as follows: Based on the three-dimensional kinematic equations of a ship, a ship dynamics model is constructed, taking longitudinal motion as an example: ; in, Expressed as the ship's actual measured full-load mass; Expressed as the standard value of seawater density; It is expressed as the waterline area, calculated from ship design drawings and draft; This is expressed as the longitudinal drag coefficient, a parameter to be calibrated. This is represented as the control cycle.

[0029] The equations for lateral motion and heading angle motion are similar, where the lateral drag coefficient... For the parameter to be calibrated, the measured calculated value of the moment of inertia is: .

[0030] In this embodiment, the parameter calibration method adopts the least squares method, and the specific steps are as follows: Step 1: Discretization and Vector Construction: Discretize the longitudinal motion equations according to the control period. Discretization yields: ; Rearranged into linear form: ; where the output vector ; The input vector is: ; because , All contain Parameter vector to be calibrated .

[0031] Step 2: Least squares solution and cross-validation, specifically as follows: Divide the 776,000 data sets into 620,800 training sets and 155,200 test sets in an 8:2 ratio; solve using the least squares method. The calibration parameters are obtained; Regularization (regularization coefficient) To avoid overfitting, the final calibration results are shown in Table 2 below: Table 2

[0032] In Table 2 above, RMSE represents the root mean square error, which is the industry's required accuracy for ship dynamics model predictions. The result meets the requirements.

[0033] In this embodiment, the improved Kalman filter is defined as: State vector: Among them, the 6 dimensions correspond to the ship's position, speed, and attitude; State transition matrix: ; Observation matrix: dynamically adjusted based on the status of the positioning sensors; under normal conditions, observation... ; The integral of the observed velocity when failure occurs.

[0034] In this embodiment, the noise covariance matrix is ​​trained using a grid search method, and the specific steps are as follows: Step 1: Observe the noise covariance matrix Initialization based on sensor nominal accuracy: Among them, GNSS position error To variance IMU speed error To variance wait.

[0035] Process noise covariance matrix : ; Step 2: Minimize the root mean square error (RMSE) between the filtered output position and the GNSS measured position. Search scope: Diagonal elements: range Step length ; Diagonal elements: range Step length The optimal parameter results are shown in Table 3 below: Table 3

[0036] The initial parameters in Table 3 above have been verified as optimal through previous experiments, so there is no change after optimization; if the initial parameters deviate, this step will significantly reduce RMSE.

[0037] Step 3: The ship remains stationary under low disturbance conditions. Data was collected continuously for 1 hour; the results are shown in Table 4 below: Table 4

[0038] In this embodiment, the specific details for solving the sensor synchronization error problem are as follows: Resolving the issue of timestamps between IMU and GNSS The specific solution to the problem of increasing long-term cumulative deviation between velocity integral position and GNSS position due to the first-order deviation is as follows: In the data preprocessing stage, a fine-grained timestamp alignment algorithm is used to control the time synchronization error within a certain range. Within, eliminate cumulative deviation.

[0039] In this embodiment, the specific steps to solve the overfitting problem of the ship dynamics model are as follows: Solving the initial calibration... At that time, the training set RMSE was However, the RMSE on the test set suddenly increased to The specific solution to the (overfitting) problem is: introduce... Regularization term (regularization coefficient) ), and restrictions The search range was [0.7, 1.0], and the final test set RMSE decreased to .

[0040] In this embodiment, the specific details regarding the solution to the robustness problem of Kalman filtering during brief GNSS failures are as follows: Solving the problem of GNSS interruption due to obstruction At that moment, the filter position deviation increased instantaneously to The specific solution to the above problem is to add adaptive noise adjustment logic to the improved Kalman filter: when GNSS data is interrupted, temporarily increase the process noise covariance. The variance of the midpoint component, i.e., from Adjusted to At the same time, the weighting of the IMU velocity integral is strengthened to control the position deviation during the interruption period within a certain range. Within.

[0041] The following core parameters are ultimately output in this embodiment and directly used in the intelligent electronic anchoring algorithm of Embodiment 2, as shown in Table 5 below: Table 5

[0042] This embodiment accurately obtained the core parameters of ship dynamics through 80 hours of real ship testing, acquisition of 776,000 sets of valid data, least squares parameter calibration, grid search optimization and noise filtering, and other research and development steps. , (and the noise covariance matrix of the improved Kalman filter).

[0043] Example 2 This embodiment is based on the ship dynamics parameters obtained from the training in Embodiment 1. , ) and the improved Kalman filter noise covariance matrix, in wind speed Water flow velocity In inland sea anchoring scenarios, the anchoring radius... center .

[0044] The intelligent electronic anchoring control method based on ships provided in this application includes: In this embodiment, the operating system is Ubuntu 20.04LTS with real-time scheduling enabled and the process priority is set to 99; the MPC solver is OSQP 0.6.2, and `eps_abs=1e4` and `eps_rel=1e4` are configured to ensure the solution accuracy. Filtering Algorithm: Improved Kalman Filter, loaded with the training from Example 1 ; ; Ship dynamics model: Embedded in Example 1 calibration , CAN bus (propeller control, baud rate 500kbps), Ethernet (ship end status monitoring).

[0045] In this embodiment, the environmental parameters are shown in Table 6 below: Table 6

[0046] In this embodiment, the control parameters are specifically shown in Table 7 below: Table 7

[0047] Based on the data in the table above, the detailed implementation steps are as follows: In this embodiment, step S1: Anchoring function activation and parameter initialization; Input: Ship end control software "charter mode" activation command; The specific steps are as follows: S11, The intelligent control module reads the preset anchoring parameters from the parameter storage unit: Anchoring radius Allowable deviation threshold ; MPC initial parameters: prediction step size Location weight matrix Control weight matrix Energy consumption weight ; S12. The GNSS receiver output of the positioning and environmental perception module activates the ship's position at that moment. Use this as the center to determine the initial anchoring target area; Output: the initial anchoring target area, i.e., the center of the circle. ,radius The system then enters the control loop.

[0048] In this embodiment, step S2: Input the raw sensor data, i.e., the acquisition frequency. (Obstacle data) ); The specific steps are as follows: S21, Data Acquisition, i.e. At any given time, the detailed sensor data collected and the valid data after filtering by "-" are shown in Table 8 below. Obstacle distances are marked as "-" because there are no obstacles, and have been automatically filtered. Table 8

[0049] S22. Redundancy fusion employs an improved Kalman filter. The specific steps are as follows: GNSS data update interval If so, it is considered normal; the observation matrix is ​​set to... ; The fused output state vector is: ; S23, Output Frequency: Fusion data... Synchronize to steps S3 and S6; output the fused state vector. .

[0050] In this embodiment, step S3: Adaptive adjustment of MPC parameters under dynamic environment: input the output of step S2 , Historical interference level This is the initial time; the specific steps are as follows: S31. Calculation of Environmental Interference Level: Substitute into the formula ,have to: It was determined to be a medium-level interference, i.e. ; S32, MPC parameter adjustment: Position weight matrix under medium interference. ; Energy consumption weight Predicted step size ; S33, Parameter Integration: Adjusting the parameters... , and , Integration; output adjusted MPC parameters, i.e. , , , .

[0051] In this embodiment, step S4: Input the fused data from step S2. The adjusted MPC parameters in step S3; the specific steps are as follows: S41, Constructing the ship dynamics model: Combining the parameters of a small ship, i.e. , , , , , ), calculate the wind / water flow forces: Wind force (longitudinal): Wind force (lateral): Forces exerted by water flow (longitudinal): Water flow force (lateral): Substitute into the longitudinal position prediction equation Lateral position prediction equation This allows us to obtain predictions of the state at future moments. S42. Construct the optimization objective function: ; in, for Always Predicted location at time for Momentary thruster control quantity (including) ); S43. Solving for control variables: Constraints: , Solved using the OSQP solver (time consumption) Solving for the current moment, we obtain the thruster control input: The comparison of this control variable under different interference scenarios; Output thruster control quantity .

[0052] In this embodiment, step S5: Input the output of step S4 The specific steps are as follows: S51, Conventional Thrust Distribution: Bow thruster: responsible for Thrust direction and Consistent; Stern propeller: responsible for Thrust direction and Consistent; Left thruster: bears the responsibility Thrust direction and Vertically to the left; Right-side thruster: responsible for Thrust direction and Vertically to the right; like Figure 2 As shown, the thrust distribution ratio of each propeller (calculated in absolute value) intuitively reflects the dominant role of the bow and stern propellers in longitudinal thrust. S52. Fault detection: Monitor the thruster load and ensure that the current is ≤120% of the rated current. If there is speed feedback within 300ms after the command is issued, it is determined that there is no fault. S53, Thrust Speed: (According to...) Convert to speed command: bow thruster Stern propulsion Left thruster Right-side thruster Output thruster speed command (the speed of each thruster is as above).

[0053] In this embodiment, step S6: Input the actual position of the ship after step S5. ,Right now The specific steps are as follows: S61. Calculate the deviation using the Euclidean distance formula: ; S62, Dynamic Branching: Because The vessel was determined to be within permissible limits; the current level of control was maintained. Return to step S2 to continue data collection; continuously monitor the anchoring deviation trend within 10 seconds. It can be seen that the deviation remains stable within the allowable range, with no risk of exceeding tolerance; maintain the output control quantity. Return to step S2.

[0054] In this embodiment, step S7: Input the running status of steps S2-S6, that is... ; The specific steps are as follows: S71, Sensor data validity: GNSS data update interval Uninterrupted data transmission from IMU and wind / water flow sensors; time consumed in solving for control inputs: Thruster status: Load current is normal, speed feedback is timely; no fault is detected. S72. Emergency Response: No faults, no alarms or backup strategies are triggered; S73. Log Management: Record key data at the moment of failure, including the fusion data from step S2. Step S4 Step S6 Stored in memory, where backup time can be up to [time period]. The logs are uploaded to the Bubble Pig server via a 4G module; the output system runs continuously and stably, with logs backed up locally and then uploaded.

[0055] In this embodiment, the issue of GNSS (10Hz) and IMU (100Hz) timestamps is resolved. The specific solution to the problem of increased cumulative position error during data fusion due to level deviation is as follows: Use a cross-correlation-based timestamp alignment algorithm to interpolate IMU data onto GNSS timestamps, controlling the synchronization error within a certain range. Internally, a time synchronization residual compensation term was added to the improved Kalman filter to further eliminate the impact of synchronization error on fusion accuracy. After verification, the RMSE of the synchronized position fusion improved from... Down to .

[0056] In this embodiment, the specific solution to the oscillation problem using 4.2MPC is as follows: Addressing the control quantity output by the MPC solver under medium to high disturbance conditions. Fluctuations exceeding The specific solution to the problem of frequent acceleration and deceleration of the thruster, increased energy consumption, and decreased anchoring accuracy is as follows: Adjust the OSQP solver parameters, changing `eps_abs` and `eps_rel` from the default values. Down to To improve the accuracy of the solution, a penalty weight for the rate of change of the control variable is added to the objective function, that is... Under high disturbance conditions, the value was adjusted from 0.3 to 0.5 to smooth control fluctuations. After verification, the fluctuation amplitude decreased from... Down to Within.

[0057] In this embodiment, as Figure 3 As shown, the specific issues regarding the verification of thruster failure response are as follows: The specific solution to address the simulated "right thruster failure (300ms no speed feedback)" and verify fault robustness is as follows: the thrust distribution unit automatically adjusts the thrust of the right thruster... All thrust is distributed to the left thruster, which bears the load. Thrust; the emergency response module triggered a "single thruster failure" alarm, and the log recorded the data at the moment of the failure; after verification, the anchoring deviation was from Slightly increased to But still ≤ The system continues to operate stably.

[0058] The output results of this embodiment verify the effectiveness of the pre-trained MPC model and the electronic anchoring algorithm through the complete control loop, as shown in Table 9 below: Table 9

[0059] This embodiment is based on the pre-trained model of Embodiment 1, and covers everything from hardware deployment and parameter initialization to multi-source data fusion, MPC control law solving, thrust allocation and monitoring adjustment, and also includes... Figures 2 to 4It intuitively displays the logic and data results of key technical links, and solves problems such as "sensor synchronization, MPC oscillation, and thruster failure".

[0060] Example 3 This embodiment is based on the schemes of Embodiment 1 and Embodiment 2. Through 120 hours of field experiments and 90 sets of independent tests, each set lasting 1200 seconds, the system performance was verified in three interference scenarios: low, medium, and high. The system performance was also quantitatively compared with existing technologies.

[0061] The following section will elaborate on the specific implementation process of this scheme using concrete experimental data and present the corresponding experimental results, providing data support and practical basis for the practical application of this scheme.

[0062] It should be noted that the experimental verification in Example 3 was not carried out in isolation. Instead, it used the algorithm application data of Example 2 as the core benchmark and the data tables recorded in Example 2, such as "sensor acquisition data, MPC control quantity solution results, thrust distribution ratio, and anchoring deviation changes". This not only set "standard operating conditions" for Example 3, but also provided reproducible algorithm parameters and logical closed loops to ensure the consistency and reliability of the experimental verification.

[0063] Key data from Example 2, such as "anchoring deviation of 0.316m under medium interference scenario, thruster control quantity [300,-200]N, and thrust distribution ratio (i.e., bow 22%, stern 29.3%, left side 24.4%, right side 24.4%)", are directly used as benchmark parameters for Example 3 "medium interference scenario". This ensures that the comparative experiments have a unified evaluation standard and avoids comparison deviations caused by parameter differences.

[0064] In this embodiment, the experiment was conducted in a closed sea area and divided into three scenarios according to the "intensity of environmental interference". The environmental parameters, test duration, and number of groups for each scenario were clearly defined. Among them, the "medium interference scenario" (i.e., wind speed 3.0~5.0m / s, water flow 0.8~1.2m / s) is completely consistent with the application scenario of Example 2. The purpose is to verify whether the stability of the algorithm in Example 2 can be reproduced under the same scenario, and at the same time, to expand the low / high interference scenarios to verify the generalization ability of the algorithm, as shown in Table 10 below: Table 10

[0065] Specifically, each test group in Table 10 was spaced 10 minutes apart to avoid the influence of ship inertia on the next set of data; the wave height was less than 0.5m during the experiment to eliminate additional interference from waves on the anchoring.

[0066] To ensure a fair comparison, this application maintains the same core control variables as the prior art, differing only in the algorithm and system architecture. These control variables are all derived from the preset parameters of Embodiment 2, such as the anchoring radius. Allowable deviation This allows the comparative experiment of Example 3 to directly use the evaluation criteria of Example 2, avoiding comparison distortion caused by parameter differences, as shown in Table 11 below: Table 11

[0067] The rationality of these parameters has been verified in Example 2. For example, the 7m anchoring radius in the charter mode can meet the needs of multiple people's activities, and the allowable deviation of 0.8m can balance accuracy and energy consumption. These parameter settings provide a unified "performance benchmark" for the comparative experiment in Example 3, making the comparison results of different methods comparable.

[0068] In this embodiment, the evaluation indicators directly correspond to the core technical problems to be solved in this application, and the calculation method completely follows the data analysis logic of Embodiment 2, such as the Euclidean distance calculation of anchoring deviation and the power integral method of thruster energy consumption, to ensure the stability of the indicator calculation of Embodiment 3 with the experimental results of Embodiment 2, as shown in Table 12 below: Table 12

[0069] Table 12 above has verified the feasibility of these indicators in Example 2 through "deviation of 0.316m under medium disturbance and energy consumption of 1.8kWh". Example 3 extends this to multiple scenarios and multiple comparison objects to further verify the superiority of the indicators.

[0070] In this embodiment, the preparation steps before the experiment are as follows: Step 1) GNSS: Calibrate using an RTK base station to ensure positioning accuracy ≤ 0.5m, consistent with the GNSS calibration method in Example 2. After calibration, the positioning error matches the 0.21m fusion accuracy in Example 2; allow to stand still for 30 minutes, perform zero-bias calibration, and ensure attitude error ≤ 0.01°. Follow the IMU calibration process in Example 2 to ensure consistent attitude data accuracy; perform no-load testing to confirm the linear relationship between thrust and rotational speed. No error, it is exactly the same as the thruster characteristic verification in Example 2, avoiding the impact of thrust output deviation on the experiment.

[0071] The intelligent control module loads the pre-trained parameters from Example 1, i.e. , , The matrices and these parameters, as verified in Example 2, can stabilize the deviation within 0.3~0.7m, providing a reliable algorithmic basis for Example 3; Set anchoring parameters: , , , We used the same initial parameters as in Example 2 to ensure that the experimental starting point was consistent with that in Example 2.

[0072] Step 2) Start the system, and the ship is moored at... Collect initial environmental data (wind speed) Water flow The data at this stage is in the same format as the "activation initialization" data in Example 2, such as initial position deviation ≤ 0.1m; The data sampling frequency is set to 10Hz, and the storage interval is 1 second (1200 storage points per group), which is compatible with the 5Hz fusion data output frequency of Example 2 to ensure consistent data processing logic.

[0073] Run the electronic anchoring algorithm of Example 2, record the anchoring deviation and thruster power. The operating logic at this stage is exactly the same as in Example 2, such as MPC parameter adaptation (under medium interference). ), thrust distribution ratio (30% at the bow, 40% at the stern) to ensure the consistency of algorithm operation; Ensure that the deviation is stable at ≤0.8m. If it exceeds the tolerance, reinitialize. The judgment criteria are basically consistent with the "deviation ≤ε" logic in Example 2 to avoid the initial instability affecting subsequent tests.

[0074] At 600 seconds, interference is introduced via a wind tunnel / water flow generator: wind speed water flow The interference intensity is higher than that in the interference scenario of Example 2, and is used to verify the anti-interference scalability of the algorithm in Example 2; the deviation change and response time (time to recover to ≤0.8m) after interference are recorded. No sudden interference was involved in Example 2. This stage is an extended verification to verify the adaptability of the algorithm under dynamic interference.

[0075] At 800 seconds, the GNSS signal was manually cut off (simulating sensor failure) or the right thruster power was cut off (simulating thruster failure). This failure scenario is consistent with the "right thruster failure" simulation in Example 2, verifying the effectiveness of the failure robustness strategy in Example 2. The average deviation was recorded within 120 seconds after the fault. In Example 2, the deviation after the fault was ≤0.8m. In Example 3, the deviation changes of different fault types were quantified based on this to further verify the robustness.

[0076] At 1000 seconds, the GNSS / thruster signal is restored, and the time it takes for the system to recover to a stable state is recorded. The recovery process is not involved in Example 2. This stage verifies the fault recovery capability of the algorithm. After completion, the system automatically calculates the five evaluation indicators for the group and stores them in the memory. The indicator calculation logic is consistent with the data analysis method in Example 2, ensuring data comparability.

[0077] Step 3) Repeated testing in multiple scenarios: Repeat 30 sets of tests in low / medium / high interference environments, with a 10-minute interval between each set (ship stationary reset). The purpose of repeated testing is to verify the stability of the algorithm in Example 2 and avoid the randomness of a single test. After the test, take the average of the 30 sets of data for each scenario and remove outliers. Outlier determination: sets with deviation > 3m to ensure data reliability. There were no outliers in Example 2. Example 3 further verifies the stability of the algorithm through repeated testing.

[0078] The specific experimental data obtained from the above experiments are shown in Table 13 below: Table 13

[0079] In this embodiment, the performance data of this application under various environmental scenarios are as follows: The average anchoring deviation in Example 3 under medium interference scenarios was 0.58m, while that in Example 2 was 0.316m. The difference stemmed from the introduction of sudden interference in Example 3, demonstrating that the algorithm in Example 2 remains stable in multiple scenarios. Furthermore, the data from extended scenarios with low / high interference further validated the algorithm's generalization ability, as shown in Table 14 below. Table 14

[0080] In this embodiment, the prior art 1 is mainly a ship end berthing and unberthing system, which has no electronic anchoring function and relies on "predetermined stopping position and segmented control", without energy consumption optimization. Embodiment 2 has clearly defined the essential difference between the "electronic anchoring (continuous position holding)" and berthing and unberthing (one-time berthing) of this application. Embodiment 3 further quantifies the advantages through energy consumption data. Existing technology 2 is a ship-shore collaborative berthing system that relies on shore-based commands and has transmission delays. In embodiment 2, the "pure ship-side autonomous decision-making" can avoid delays, and embodiment 3 quantifies the advantages through response time. Existing technology 3 uses fixed-parameter PID control without adaptive adjustment, resulting in large deviations under high interference. In embodiment 2, the MPC parameters are adaptive (under high interference). (This is superior to fixed parameters; Example 3 was verified through multi-scenario data.) Existing technology 4 is a non-redundant sensing system with only single GNSS positioning. In case of failure, there is no compensation, and the accuracy drops sharply. In embodiment 2, the redundancy fusion of the improved Kalman filter is better than that of a single sensor. Embodiment 3 has advantages through fault deviation quantification.

[0081] The comparison data is shown in Table 15 below: Table 15

[0082] In summary, the advantage of this embodiment in the sensing layer is that the redundancy fusion accuracy far exceeds that of traditional single sensors: In Example 2, the fusion accuracy of the improved Kalman filter has been verified. When the GNSS is normal, the position RMSE is 0.21m, and when the GNSS fails, the IMU compensation deviation is ≤0.5m. In contrast, the existing single GNSS positioning RMSE is 0.32m, and the deviation increases sharply to 3.21m when the fault occurs.

[0083] Example 3 further quantifies: Under medium interference, the fault robustness deviation of this application is 0.78m, while that of the prior art is 3.21m. This proves that the redundant sensing design of Example 2 fundamentally solves the defect of "failure equals failure" of traditional single sensors. This advantage is more significant in high interference scenarios, that is, the deviation of 0.92m in this application is compared with 64.5m in the prior art.

[0084] The advantage of this embodiment in the control layer is that the MPC adaptive parameters are superior to the traditional fixed parameters: In Embodiment 2, the MPC parameters are dynamically adjusted according to the interference level, resulting in low interference... (Prioritize energy saving), high interference (Prioritize accuracy); while the existing technology of traditional PID uses a fixed proportional coefficient (such as Kp=2.0), which cannot adapt to high interference.

[0085] Data from Example 3 shows that under high interference, the deviation of this application is 0.75m, and the PID is 1.13m, with a deviation reduction of 33.6%. In terms of energy consumption, under low interference, the energy consumption of this application is 1.2kWh, and the PID is 1.8kWh, with an energy saving of 33.3%. This fully demonstrates that the MPC adaptive strategy of Example 2 is superior to the traditional fixed parameter control, achieving a dynamic balance between "accuracy and energy consumption".

[0086] The advantage of this embodiment at the execution layer is that it provides robust thrust allocation to avoid traditional thrust interruption: In Embodiment 2, a "same-direction thrust compensation" strategy is designed. When the right thruster fails, the left thruster bears 100% of the lateral thrust, and the deviation increases by only 0.2m; while the traditional system has no fault allocation logic, and the thrust is directly interrupted after the failure of a single thruster, and the deviation increases by more than 1m.

[0087] Quantitative verification in Example 3: Under moderate interference, the fault deviation of this application is 0.78m, while that of the prior art is 1.56m, a reduction of 50.0%. This proves that the thrust distribution strategy of Example 2 ensures anchoring continuity from the execution layer and avoids the defect of "stopping immediately upon failure" in traditional methods.

[0088] The advantage of this embodiment at the architecture layer is that it avoids shore-based delays through purely ship-based autonomous decision-making. Example 2 clearly states that "there is no intervention from the shore-based command, only log uploads", with a decision delay of <0.1s; existing technologies rely on shore-based data interaction, with a delay of ≥2s.

[0089] Data from Example 3 shows that the response time of this application under moderate interference is 1.5s, while that of the prior art is 3.8s, representing a 60.5% improvement in response speed. This demonstrates that the pure ship-side architecture of Example 2 fundamentally solves the problem of "delay leading to lag" in traditional ship-shore collaboration, and its advantages are particularly prominent in sudden interference scenarios.

[0090] In this embodiment, the results of Example 3 are shown in Table 16 below: Table 16

[0091] This embodiment, through 120 hours and 90 sets of reproducible experiments, using the algorithm application data of Embodiment 2 as a benchmark, comprehensively verifies the significant advantages of this application in terms of accuracy and energy consumption balance, fault robustness, and response speed.

[0092] Example 4 This embodiment describes in detail the intelligent electronic anchoring control system based on embodiments 1 and 2, and verifies it through actual ship testing. The intelligent electronic anchoring control system based on the ship consists of a positioning and environmental perception module 1, an intelligent control module 2, a power regulation and actuator 3, an emergency handling module 4, and a data interaction module 5. At the same time, it relies on the Bubble Pig software 6 and the Bubble Pig server 7 to realize remote interaction.

[0093] like Figure 4 The image shows a ship-based intelligent electronic anchoring control system provided in an embodiment of this application. The system includes: a positioning and environmental sensing module 1, responsible for collecting ship status data, environmental interference data, and obstacle data, including: Positioning sensor 11 uses a Trimble R10 GNSS receiver, supports RTK positioning, and has static accuracy. Dynamic accuracy It updates at a frequency of 10Hz and outputs the ship's real-time latitude, longitude, and altitude information as a positioning reference for anchoring the center of the circle.

[0094] Inertial measurement unit 12 uses ADIS16488 six-axis IMU, and its measurement range covers angular velocity. acceleration angular velocity accuracy , acceleration accuracy It updates at a frequency of 100Hz and outputs ship attitude (heading angle, roll angle, pitch angle) and linear velocity information to compensate for the lack of dynamic response in GNSS.

[0095] Wind direction sensor 13 uses an Aosong AWM3000 ultrasonic anemometer, with a measurement range of... Wind speed accuracy Wind direction accuracy It updates at a frequency of 5Hz, collecting real-time wind speed and direction to provide a basis for environmental disturbance assessment.

[0096] The water flow sensor 14 uses a YSI556MPS Doppler current meter, with a measurement range of... Water flow velocity accuracy Flow accuracy The system updates at a frequency of 5Hz, collects water flow velocity and direction data, and uses this data, along with wind direction data, to calculate environmental disturbance forces.

[0097] The millimeter-wave radar 15 uses the Continental ARS408 millimeter-wave radar, with a detection range of... angular resolution With a refresh rate of 10Hz, it can detect the distance and angle of obstacles around the ship in real time, ensuring safety during the anchoring process.

[0098] Specifically, vessel status data includes position. Heading angle yaw angle Pitch angle Longitudinal velocity and lateral speed ; Among them, location The heading angle is expressed as the ship's real-time position coordinates at time t, reflecting the ship's current position and status. The angle between the ship's direction of travel and true north is represented as 0°-360°, determining the direction of the propeller thrust; the roll angle... The pitch angle is expressed as the ship's inclination angle about its longitudinal axis, set at ±45°, to help correct the ship's dynamics model; Expressed as the ship's inclination angle about its transverse axis, set at ±45°, the draft is estimated to correct for the drag coefficient; longitudinal speed Expressed as the ship's speed along the course, used to calculate the predicted position at the next moment; lateral speed. Expressed as the ship's speed perpendicular to its course, compensating for lateral drift; environmental disturbance data includes wind speed W and wind direction. Water flow velocity C, water flow direction Among them, environmental interference data includes wind speed W, which represents the airflow velocity, used to calculate the force of wind on the ship; wind direction; and wind direction. The angle between the wind's direction of travel and true north is represented as 0°–360°, determining the direction of the wind force; the water flow velocity C represents the water flow speed, used to calculate the force exerted by the water flow on the ship; the water flow direction... It is represented as the angle between the direction of water flow and due north, set at 0°-360°, to determine the direction of the force exerted by the water flow; Obstacle data refers to the distance information of obstacles within 1m to 30m around the ship, reflecting the relative position of the ship and surrounding obstacles, and ensuring that the initial anchoring target area does not contain obstacles.

[0099] Intelligent control module 2, acting as the system's "brain," is responsible for parameter calling, data fusion, parameter adjustment, and control quantity calculation, including: The hardware carrier of the MPC control law solving unit 21 is an Advantech UNO2484G embedded industrial computer (CPU is Intel Celeron J4125, 4 cores 2.0GHz; memory 8GB DDR4; storage 128GB SSD, supports a wide temperature range of 40~85℃), integrating the OSQP0.6.2 quadratic programming solver (compiled as a C++ dynamic library, encapsulated and called through Python, with a solution accuracy of...). , Based on the ship dynamics model and environmental data, the propeller control quantities that meet the accuracy and energy consumption constraints are solved.

[0100] The redundant filtering unit 22 deploys an improved Kalman filter algorithm to fuse the data from the positioning sensor 11 and the inertial measurement unit 12; when the GNSS signal is blocked, the position deviation is compensated by integrating the IMU data to ensure the continuity and accuracy of positioning.

[0101] The adaptive weight adjustment unit 23 dynamically adjusts parameters such as position weight and energy consumption weight in the MPC control law based on the environmental interference level output by the redundant filtering unit 22 (calculated from the data of the wind direction sensor 13 and the water flow sensor 14), so as to achieve a balance between "precision and energy consumption" under different interference scenarios.

[0102] Parameter storage unit 24 pre-stores the initial parameter mapping table (such as anchoring radius) of step S1 in embodiment 1. Allowable deviation MPC prediction step size etc.), and the ship dynamics model parameters calibrated in Example 1 (etc.) , This provides the basic parameters for solving the MPC control law.

[0103] The power regulation and actuator 3 module is responsible for thrust distribution and command execution, converting the control output from the intelligent control module 2 into thruster actions, including: Electric thrusters 31, equipped with four Torqeedo Cruise 10.0R electric thrusters, divided into bow thruster 311, stern thruster 312, port thruster 313, and starboard thruster 314; each unit has a rated thrust of 500N, power of 2.4kW, and a speed range of 0-2000r / min, with a linear relationship between thrust and speed. (This provides power for anchoring ships.)

[0104] The thrust distribution unit 32 uses a self-developed CAN bus distribution board (STM32F407 main controller) to receive the total thrust control quantity output by the MPC control law solving unit 21, and distribute it to each electric thruster 31 according to a preset ratio (30% at the bow, 40% at the stern, 50% on the left, and 50% on the right) to generate thrust commands for each thruster.

[0105] The thruster drive unit 33 uses the original thruster drive module (TorqeedoPower485000), with an input voltage of 48V and a maximum current of 100A. After receiving the command from the thrust distribution unit 32, it controls the rotation speed of the electric thruster 31 in a closed loop to ensure that the actual thrust deviates from the command by ≤2%.

[0106] Emergency Response Module 4 is responsible for fault detection, emergency stop response, and log management to ensure safe operation of the system during abnormal situations. This includes: The fault detection unit 41 uses a self-developed sensor / thruster status monitoring board (STM32F103 main controller) to monitor the signal integrity of the positioning sensor 11 and the inertial measurement unit 12, as well as the current, voltage and temperature of the electric thruster 31 at a frequency of 100Hz. When the GNSS interruption is greater than 1s or the thruster current exceeds the rated value by 120%, it is determined to be a fault.

[0107] The emergency stop response unit 42 integrates an electromagnetic brake with a coil voltage of 24V, a response time of ≤200ms, and a braking force of 1000N. When the fault detection unit 41 determines a serious fault (such as failure of the dual thrusters), it is immediately triggered to ensure that the ship stops moving within 2s.

[0108] The log recording unit 43 is equipped with a Class 10 memory (32GB) to record key data such as fault codes, anchoring deviation, and thruster power at a frequency of 1Hz, with a cyclic storage time of 72 hours, providing a basis for fault backtracking and performance analysis.

[0109] The audible and visual alarm 44 uses a 24V audible and visual alarm. When a fault occurs, the light flashes at a frequency of 2Hz and the volume reaches 105dB, ensuring that it can be identified within 50m and reminding operators to deal with it in time.

[0110] Data interaction module 5 is responsible for function activation, log uploading, and anchoring status synchronization to achieve ship-shore interaction, including: The Ethernet module 51 uses an industrial-grade Ethernet module (WIZnetW5500), supporting TCP / IP communication at a speed of 100Mbps; Figure 4 As shown, it receives instructions such as "Electronic Anchoring Function Activation" sent by the shore-based Bubble Dolphin software 6 and forwards them to the intelligent control module 2.

[0111] The 4G module 52 uses the Huawei ME909u521 module, supports LTE Cat4, and has an upload speed of 150Mbps. It uploads the data from the log recording unit 43 to the Bubble Pig server 7 at 10-second intervals, and synchronizes the anchoring status every 1 second.

[0112] Bubble Pig Software 6 runs on a shore-based computer, providing a human-computer interaction interface. It supports sending "electronic anchoring function activation" commands, configuring initial parameters such as the anchoring center, and viewing real-time anchoring status and historical logs.

[0113] Bubble Dolphin Server 7 is deployed on a cloud server, receiving log data uploaded by 4G module 52, storing, analyzing and visualizing it, providing data support for multi-ship anchoring management and algorithm iteration.

[0114] In this embodiment, the modules are connected via CAN bus, Ethernet, and GPIO to form a closed loop of "sensing, computing, control, execution, and monitoring": The raw data from positioning sensor 11 (NMEA0183 protocol), inertial measurement unit 12, wind direction sensor 13 (Modbus RTU protocol), water flow sensor 14 (Modbus RTU protocol), and millimeter-wave radar 15 (CAN bus custom protocol) are transmitted to intelligent control module 2 every 10ms via RS232 / RS485 to CAN module at a baud rate of 500kbps.

[0115] The thruster control quantity generated by the MPC control law solving unit 21 is converted into a single thruster command by the thrust distribution unit 32 and sent to the thruster drive unit 33 every 100ms via the CAN bus at a baud rate of 500kbps.

[0116] The current, voltage, and temperature status of the electric thruster 31 (CAN bus custom protocol) are synchronized to the fault detection unit 41 every 100ms via the thruster drive unit 33.

[0117] The intelligent control module 2 receives instructions from the Bubble Pig software 6 via the Ethernet module 51; the log recording unit 43 of the emergency processing module 4 uploads data via the 4G module 52, and the anchoring status is synchronized to the Bubble Pig server 7 every 1 second.

[0118] When the fault detection unit 41 determines a serious fault, it triggers the emergency stop response unit 42 through a GPIO high-level signal, and at the same time sends an emergency stop command to the intelligent control module 2 and the power adjustment and actuator 3.

[0119] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0120] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Although embodiments of this application have been shown and described above, it is understood that the above embodiments should not be construed as limitations on this application, and those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A ship-based intelligent electronic anchoring control method, characterized in that, Includes the following steps: S1. Activate the electronic anchoring function and obtain the matching initial anchoring parameters and initial anchoring target area according to the current operating mode. The initial anchoring parameters include the anchoring radius R, the allowable deviation threshold ε, and the MPC initial parameters. Based on the initial anchoring target area, the target area is centered on the ship's position at the moment of activation. ; S2. Collect vessel status data, environmental interference data, and obstacle data based on the initial anchoring target area, perform redundant fusion processing on the collected multi-source data, and output fused data. S3. Calculate the environmental interference level based on the environmental interference data in the fused data. According to the level of environmental interference Dynamically adjust the initial MPC parameters in step S1 and output the adjusted MPC parameters; S4. Based on the ship state data and adjusted MPC parameters from the fused data, construct an optimization objective function and a ship dynamics model with energy consumption constraints, and solve for the propeller control variables. ; S5, based on the thruster control quantity The thrust is distributed to each electric thruster according to a preset rule. When a thruster malfunction is detected, the dynamic adjustment of the distribution ratio is triggered, and a thrust command is output. S6. Collect the actual position of the ship after executing the thrust command. Calculate the actual position of the ship Center of the target area deviation According to the deviation The logic triggers a return to step S2 or step S3; S7. Monitor the running status of steps S2-S6 in real time. When a fault is detected, trigger an alarm and execute the backup strategy. At the same time, record and upload the fault data to form a closed-loop control.

2. The intelligent electronic anchoring control method based on ships according to claim 1, characterized in that, In step S2, based on the initial anchoring target area determined in step S1, the positioning and environmental perception module (1) is deployed. The positioning and environmental perception module (1) includes a positioning sensor, an inertial measurement unit, a wind direction sensor, a water flow sensor, and an obstacle detection sensor. The specific steps are as follows: S21, Positioning and Environmental Perception Module (1) collects ship status data, environmental interference data, and obstacle data at a preset frequency; Ship status data includes position Heading angle yaw angle Pitch angle Longitudinal velocity and lateral speed ; Environmental disturbance data include wind speed (W) and wind direction. Water flow velocity C, water flow direction ; Obstacle data refers to the distance information of obstacles within 1m to 30m around the ship; S22. The data acquired in step S21 is fused using an improved Kalman filter algorithm. The state vector of the improved Kalman filter algorithm is defined as follows: ； in, For the ship's position, These are the longitudinal and lateral velocities, respectively. These are the pan and tilt angles, respectively. The observation matrix of the improved Kalman filter algorithm is Observation matrix The state vector The linear mapping is the predicted observation vector at the current time. The status of the positioning sensor is obtained. If the data update interval of the positioning sensor is ≤0.1s, it is considered normal; if the data interruption of the positioning sensor is >5s, it is considered faulty. The observation matrix is ​​dynamically adjusted based on the status of the positioning sensors. : When the positioning sensor determines that it is normal, then the observation matrix... Set as: , observation ; When the positioning sensor determines that it has failed, the observation matrix... Set as: Velocity integral completion via inertial measurement unit ; S23. The data fused in step S22 is output to steps S3 and S6 respectively at a preset frequency.

3. The intelligent electronic anchoring control method based on ships according to claim 2, characterized in that, The initial parameters for MPC in step S3 include the initial prediction step size. Initial position weight matrix Initial control weight matrix and initial energy consumption weight The specific steps of step S3 are as follows: S31. Based on the wind speed W and water flow velocity C in the environmental interference data output in step S2, introduce the historical interference attenuation factor. Calculate the current environmental interference level : ； in, The interference level at the previous moment. , ; S32. Environmental interference level calculated based on step S31 Adjust the position weight matrix Q and energy consumption weight in the initial parameters of MPC. : like If so, it is determined to be low interference: , ; in, This is the initial position weight matrix in step S1. As the initial energy consumption weight; like If so, it is determined to be a medium-level interference: , ; like If so, it is determined to be high interference: , ; S33, Adjust Q after step S32. and the initial prediction step size in step S1 and the initial control weight matrix Perform the integration operation to obtain the adjusted MPC parameters and output them to step S4.

4. The intelligent electronic anchoring control method based on ships according to claim 3, characterized in that, In step S4, based on the vessel state data and environmental disturbance data output in step S2, and combined with the inherent parameters of the small vessel, the following three-dimensional dynamic model is constructed: Longitudinal position prediction equation: ； in, t represents the control period; t represents the current control time; k represents the prediction step index of the MPC algorithm. Let be the longitudinal position coordinates of the ship at time t+k; Let be the longitudinal velocity of the ship at time t+k; Let be the total longitudinal thrust of the ship at time t+k; This refers to the longitudinal force exerted by the wind on the ship. ρ represents the longitudinal force exerted by the water flow on the ship; m represents the mass of the ship. The predicted longitudinal position of the ship at time t+k+1; Lateral position prediction equation: ； in, Let be the lateral position coordinates of the ship at time t+k; Let be the lateral velocity of the ship at time t+k; Let be the total lateral thrust of the ship at time t+k; The lateral force exerted by the wind on the ship; The lateral force exerted by the water flow on the ship; The predicted lateral position of the ship at time t+k+1; and ; ; Heading angle prediction equation: ； in, Let be the heading angle of the ship at time t+k; Let be the angular velocity of the ship at time t+k; Let t+k be the bow turning moment of the ship; The turning moment of the ship due to wind; I is the turning torque of the water flow on the ship; I is the moment of inertia of the ship about its vertical axis. The predicted heading angle of the ship at time t+k+1; and ; ; Bow torque equation: ； in, For the turning torque, Let A be the density of seawater and A be the surface area of ​​the waterline. , These are the longitudinal and lateral drag coefficients, respectively. , These are the thrusters at the stern and bow, respectively. , B is the distance from the propeller to the center of the ship, and B is the width of the ship. , These are the thrusters on the right and left sides of the ship, respectively.

5. The intelligent electronic anchoring control method based on ships according to claim 4, characterized in that, Step S4 also includes the following specific steps: S41. Construct the optimization objective function based on the adjusted MPC parameters: in, For time t Predicted location at time The initial anchoring target area center in step S1, for The generalized control input of the time-of-flight thruster is defined by N, where N is the MPC prediction step size, and Q and R are the position weight matrix and control input weight matrix, respectively. Energy consumption weighting coefficient This is an energy consumption constraint term; S42. Constrain the amplitude and anchoring range of the control quantity. Solve the optimization objective function of step S41 using a quadratic programming algorithm to obtain the thruster control quantity at the current moment. And output it to step S5; The control amplitude constraint is as follows: ,in For the minimum value of the thruster control quantity, Set to the maximum value of the thruster control quantity; The anchoring range constraint is: , where R is the anchoring radius determined in step S1.

6. The intelligent electronic anchoring control method based on ships according to claim 5, characterized in that, The specific steps of step S5 are as follows: S51. Output from step S4 The electric propulsion units, including the bow propulsion unit, stern propulsion unit, port propulsion unit, and starboard propulsion unit, are allocated according to the following rules: Bow thrusters: The thrust direction and the heading angle output in step S2 Consistent; Stern thruster: Thrust direction and Consistent; Left thruster: Thrust direction and Vertically to the left; Right-side thruster: Thrust direction and Vertically to the right; S52. Obtain the load and response status of each thruster; When a single thruster fails, it triggers an adjustment to the lateral thrust distribution ratio; If both thrusters fail, longitudinal thrust is retained, triggering the emergency handling in step S7. S53. Convert the thrust of each propeller determined in step S51 or S52 into a corresponding speed command, output it to the propeller for execution, and after execution, adjust the ship's position and feed back the status to step S2.

7. The intelligent electronic anchoring control method based on ships according to claim 6, characterized in that, The specific steps of step S6 are as follows: S61. Collect the actual position of the ship after step S5 is executed according to the preset frequency. ,calculate With respect to the center of the initial anchored target area in step S1 Euclidean deviation ; S62, if Allowable deviation threshold If the current thruster control value is maintained, return to step S2 to continue data collection; if Allowable deviation threshold Then Feedback is sent to step S3, and steps S3-S5 of MPC parameter adjustment and thrust control are re-executed.

8. The intelligent electronic anchoring control method based on ships according to claim 7, characterized in that, The specific steps of step S7 are as follows: S71. Monitor the sensor data in step S2, the time taken to solve the control quantity in step S4, and the thruster fault status in step S5 in real time. If the sensor data interruption exceeds 5 seconds, the time taken to solve the control quantity exceeds 1 second, or the thruster fails, it is determined to be a fault. S72. Trigger an audible and visual alarm when a fault occurs. If a sensor fault is confirmed, enable the redundant fusion data in step S2. If a thruster malfunction is determined, the adjustment strategy in step S52 is executed; If a serious malfunction is identified, all thruster thrust is cut off and mechanical braking is activated. S73. Record key data at the time of the fault. Key data includes the fused data from step S2 and the thruster control quantity from step S4. Deviation in step S6 Key data is backed up locally and then uploaded to the server.

9. A system for implementing the ship-based intelligent electronic anchoring control method according to any one of claims 1 to 7, characterized in that, include: The positioning and environmental perception module (1) is used to perform multi-source data acquisition in step S2 of claim 1, including a positioning sensor, an inertial measurement unit, a wind direction sensor, a water flow sensor and an obstacle detection unit, and outputs raw acquisition data; The intelligent control module (2) is used to perform parameter calling in step S1 of claim 1, redundancy fusion in step S2, parameter adjustment in step S3, and control quantity solving in step S4. It includes an MPC control law solving unit, a redundancy filtering unit, an adaptive weight adjustment unit, and a parameter storage unit. The power adjustment and execution mechanism (3) is used to perform the thrust distribution and command execution of step S5 of claim 1, and includes at least a plurality of electric thrusters, a thrust distribution unit and a thruster drive unit; Emergency handling module (4) is used to perform fault handling and log management in step S7 of claim 1, and includes at least a fault detection unit, an emergency stop response unit, a log recording unit and an alarm unit; The data interaction module (5) is used to realize the function activation in step S1, the log upload in step S7 and the anchoring status synchronization, and includes at least an Ethernet unit and a wireless communication unit. The positioning and environmental perception module (1), intelligent control module (2), power regulation and actuator (3), emergency handling module (4) and data interaction module (5) are connected in sequence to form a closed-loop system of perception, calculation, control, execution and monitoring.

10. The system of the intelligent electronic anchoring control method based on ships according to claim 9, characterized in that, The positioning sensor in the positioning and environmental perception module (1) is a GNSS receiver; the inertial measurement unit is a six-axis IMU; and the obstacle detection unit is a millimeter-wave radar. The hardware carrier of the intelligent control module (2) is an embedded industrial computer, which pre-stores the initial parameter mapping table of step S1 as claimed in claim 1 and the ship dynamics model parameters of step S4, and integrates the OSQP quadratic programming solver in the software. The emergency response unit in the emergency handling module (4) integrates an electromagnetic brake to stop the ship within 2 seconds.

Images