Adaptive wave disturbance control method for unmanned ships based on digital twin

Abstract

The present invention provides an adaptive wave interference control method for unmanned ships based on digital twins. Wave information is acquired in real time through sensors, the main frequencies are extracted by frequency domain analysis, and the LSTM model is used to predict the future wave state by combining the filtered signal and physical characteristics. The speed and heading of the unmanned ship are adjusted in real time through the ROS node to dynamically respond to changes in wave height, speed and direction to ensure stable navigation. A high-dimensional Kalman filter is used to realize virtual-real synchronous feedback and correct the control strategy. The system includes wave information acquisition, data processing and prediction, control decision-making, and feedback and optimization modules to ensure the stable navigation of the unmanned ship in a wave mutation environment.

Description

Technical Field

[0001] The present application belongs to the field of unmanned ship control technology, and specifically relates to an unmanned ship adaptive wave interference control method based on digital twin. Background Art

[0002] An unmanned ship is a surface robot that can sail autonomously on the water according to preset tasks without remote control, with the help of precise satellite positioning and its own sensors. This "surface robot" integrates multiple technologies such as ships, communications, automatic control, remote monitoring and networked systems, and realizes multiple functions such as autonomous navigation, intelligent obstacle avoidance, long-distance communication, real-time video transmission and networked monitoring. As a new technical means, unmanned ships have been widely used in the fields of marine surveying and maritime defense.

[0003] During the navigation of unmanned ships, sudden wave changes caused by wind, waves and other factors are often encountered. In this case, the rapid changes in wave height, speed and direction pose a great challenge to the stability of unmanned ships, and unmanned ships are prone to capsizing, stalling or violent swaying. Therefore, effective control of unmanned ships in wave mutation and high-dynamic environments is directly related to the driving safety of unmanned ships and the successful completion of tasks. The current unmanned ship control method can ensure smooth navigation in low sea conditions, but in complex environments with sudden wave changes, the existing control method cannot make rapid adaptive adjustments to the unmanned ship according to the real-time sea conditions, and it is difficult to achieve rapid and accurate adjustment of the unmanned ship's heading and speed, and thus cannot ensure the smooth driving of the unmanned ship under sudden wave changes.

[0004] Therefore, in view of the instability of unmanned ship navigation in a sudden wave environment and the limitations of existing control methods, a control method that can sense the changes in sea conditions in real time and adaptively adjust the heading and speed of the unmanned ship is urgently needed. Such a method can ensure that the unmanned ship can achieve stable navigation in a complex environment with sudden wave changes, ensuring its driving safety and the smooth completion of the mission. Summary of the invention

[0005] Based on the above problems, the present application provides an adaptive wave interference control method for an unmanned ship based on digital twin, and its technical solution is:

[0006] An adaptive wave interference control method for an unmanned ship based on digital twins comprises the following steps:

[0007] S1. Collect wave signals through wave radar sensors to obtain time domain information of wave height, wavelength, wave speed and direction;

[0008] S2. Preprocess the collected wave signals, remove noise and extract the main frequency and dominant period features to form a wave feature matrix X , to reflect the dynamic characteristics of waves;

[0009] S3. The wave feature matrix X Input the LSTM model and use historical wave data for training to learn the characteristics and changing trends in the wave time series data and predict the height, speed and direction of future waves. By optimizing the model's loss function, minimize the error between the predicted value and the actual value and continuously adjust the model parameters.

[0010] S4. According to the predicted future wave state, the controller obtains the wave prediction results and adjusts the heading and speed of the unmanned ship in real time; when the wave height, speed or direction exceeds the set threshold, the adjusted control command is dynamically generated to make the unmanned ship adaptively adjust and maintain stable navigation;

[0011] S5. Use a high-dimensional Kalman filter to feed back the actual state of the unmanned ship to the virtual model, optimize the control strategy based on the feedback information, ensure the synchronization of the virtual and real states, and ensure that the unmanned ship maintains stability and navigation efficiency under wave interference.

[0012] Preferably, in step S2:

[0013] S21. Using generalized Fourier transform to analyze wave signals Perform frequency domain analysis and convert it into wave spectrum in the frequency domain ;

[0014] S22. Extracting the main frequencies through spectrum analysis , which is the frequency with the most concentrated energy in the spectrum; first, in the spectrum Find the frequency range where the energy distribution is concentrated , and then determine the main frequency by weighted averaging:

[0015] ;

[0016] in, is the energy of the corresponding frequency, is a frequency variable Minimum value, is a frequency variable Maximum value.

[0017] Preferably, in step S2, the main frequency extracted is calculated The reciprocal of gets the dominant period of the wave , and calculate the response time window of the unmanned ship based on the dominant cycle ,This response time window determines the shortest time for the unmanned ship to complete the adjustment before the wave impact.

[0018] Preferably, the training process of the LSTM model includes:

[0019] S3.1 The wave characteristic matrix Input LSTM model, where is the filtered sensor time domain signal, It represents the characteristic parameters extracted from the physical model, including the main frequency and the dominant period, describing the dynamic characteristics of the wave;

[0020] S3.2 Processing the input feature matrix through multi-layer LSTM units , calculate the hidden state , recursively capture the wave timing characteristics, and use bidirectional LSTM to process the sequence data forward and backward to obtain the complete hidden state;

[0021] S3.3 will hide the state Input to the fully connected layer to generate predictions for future wave conditions , including the height, speed and direction of the waves;

[0022] S3.4 Calculate the predicted value The actual wave state The error between them is used to optimize the model by minimizing the loss function.

[0023] S3.5 uses the back propagation algorithm to adjust the model parameters and perform gradient descent optimization on the weight matrix.

[0024] Preferably, in step S4:

[0025] S41. Use LSTM model to calculate future wave state ,The controller issues wave prediction instructions to each control node in the form of ROS messages for control adjustment of the unmanned ship;

[0026] S42. The control node obtains the wave prediction instruction and adaptively adjusts the heading and speed of the unmanned ship according to the wave prediction result;

[0027] S43. The unmanned ship receives the heading and speed Instructions to dynamically adjust the rudder angle and propeller speed of the ship.

[0028] Preferably, in step S4, the unmanned boat is adaptively adjusted as follows:

[0029] Condition 1: When the wave height Greater than high threshold (e.g. 1.5m), the unmanned ship should reduce its speed to reduce the impact of waves on the hull. The new speed Current speed Subtract a certain percentage:

[0030] ;

[0031] in is the deceleration factor (e.g. 0.3);

[0032] Condition 2: When the wave height is below the low threshold When the wave is smaller than 0.5m, the impact of waves on navigation is small, and the speed can be increased appropriately to speed up the journey:

[0033] ;

[0034] in is the growth factor (e.g. 0.1);

[0035] Condition 3: When the wave direction Heading with the unmanned ship The difference is greater than the set angle threshold (e.g. 30°), the new heading Adjust to be consistent with or close to the wave direction to reduce lateral forces:

[0036] ;

[0037] Condition 4: When the wave speed Exceeding speed threshold (For example, 2m / s), it means that the wave energy is high and has a greater impact on the stability of the hull. At this time, the speed should be reduced to enhance the hull control ability:

[0038] ;

[0039] in is the deceleration factor (e.g. 0.2);

[0040] Preferably, when the unmanned ship executes the control command, its real-time status is published to the control node in the form of ROS message, which constitutes the actual status of the unmanned ship. .

[0041] Preferably, the actual state of the unmanned ship is reflected by the multi-dimensional real-time feedback of the high-dimensional Kalman filter. Feedback to the virtual model to compare the virtual model state estimates and actual status , calculate the error and correct the virtual model to ensure the consistency between the virtual model and the physical unmanned ship.

[0042] Preferably, according to the real-time error and , the control node ( / wave_adaptive_control) dynamically adjusts the control strategy;

[0043] When the speed error Exceeding the speed When the deceleration factor is 15%, Increase by 0.1-0.2, reduce the speed of the unmanned ship to ensure stable navigation. When the speed error is less than 5% of the expected speed, reduce the deceleration factor or increase the acceleration factor , to increase the speed of unmanned ships and enhance navigation efficiency;

[0044] When the heading error When the heading error exceeds 15°, the sensitivity of the heading adjustment is increased, and the adjustment step is increased by 10-30%, so as to accelerate the steering of the unmanned boat to sail against the waves. When the heading error is less than 5°, the sensitivity is reduced to prevent over-adjustment and ensure the smooth sailing of the unmanned boat.

[0045] The unmanned ship adaptive wave disturbance control system based on digital twin includes:

[0046] Wave information collection module: collects data such as wave height, speed and direction in real time, and provides basic information about the sea conditions around the unmanned ship;

[0047] Data preprocessing and feature extraction module: Filter and denoise the collected wave data, extract the main frequencies and characteristic parameters, and construct the wave feature matrix for subsequent model prediction;

[0048] Wave state prediction module: Use the LSTM model to train and predict the wave feature matrix, generate prediction results of future wave states, and support real-time decision-making of unmanned ships;

[0049] Unmanned ship control decision module: According to the wave prediction results, formulate the heading and speed adjustment strategy of the unmanned ship to ensure that the unmanned ship can adapt to the wave environment and maintain stable navigation;

[0050] Virtual-reality feedback and optimization module: The actual navigation status of the unmanned ship is fed back to the digital twin model, and the virtual-reality state synchronization is achieved through a high-dimensional Kalman filter. According to the optimized control strategy, the unmanned ship's adaptive optimization control is achieved.

[0051] Compared with the prior art, the present application has the following beneficial effects:

[0052] 1. By introducing the LSTM model to train the wave feature matrix, accurate prediction of wave height, speed and direction is achieved. Compared with the traditional control method based on experience and preset rules, the present invention can predict the future wave state in real time, and dynamically adjust the heading and speed of the unmanned ship according to the prediction results, achieving more efficient adaptive control.

[0053] 2. Use filtering technology to extract key characteristic parameters such as the main frequency and dominant period in the wave signal, and combine them with the physical model characteristics to construct an input matrix to further optimize the wave prediction model. Compared with simple wave data analysis, this invention fully considers the multi-level nature of wave characteristics and improves the model's sensitivity to wave interference and prediction accuracy.

[0054] 3. A high-dimensional Kalman filter is introduced to feed back the actual state of the unmanned ship to the digital twin model, compare the error between the virtual model and the actual state, dynamically adjust the control strategy, and achieve virtual-real synchronization and self-optimization control. This closed-loop feedback mechanism improves the control accuracy and real-time response capability of the unmanned ship, allowing it to maintain stable navigation in complex sea conditions.

[0055] 4. The present invention sets thresholds for different wave heights, directions and speeds according to the real-time wave status, and dynamically generates adjustment strategies. For example, when the wave height is too high, the unmanned ship slows down to ensure stability; when the wave direction deviation is large, the course is adjusted in time to sail into the waves. This adaptive adjustment strategy is more flexible than the existing method based on fixed control parameters, ensuring that the unmanned ship can still maintain efficient navigation under changing sea conditions.

[0056] 5. By real-time monitoring of the speed error and heading error of the unmanned ship, and adjusting the deceleration coefficient, acceleration coefficient and heading sensitivity according to the error, the unmanned ship is ensured to travel at the optimal speed and heading. This improves the navigation efficiency of the unmanned ship in a wave environment and ensures navigation safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is the flow chart of this application.

[0058] Figure 2 This is a flow chart of wave processing and prediction for this application.

[0059] Figure 3 This is a flow chart of the adaptive control strategy of this application. DETAILED DESCRIPTION

[0060] The following will be combined with the embodiments of the present invention and the drawings of the specification to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] An adaptive wave interference control method for an unmanned ship based on digital twins comprises the following steps:

[0062] S1. Collect wave signals through wave radar sensors to obtain time domain information of wave height, wavelength, wave speed and direction;

[0063] S11. Use wave radar sensor to collect wave signals in real time , including wave height ,wavelength , wave speed and wave direction The time domain information, this time domain signal Reflects the changes of waves in the time dimension:

[0064] .

[0065] S2. Preprocess the collected wave signals, remove noise and extract features such as main frequency and dominant period to form a wave feature matrix X , to reflect the dynamic characteristics of waves;

[0066] S21. Using generalized Fourier transform to analyze wave signals Perform frequency domain analysis and convert it into wave spectrum in the frequency domain :

[0067] ;

[0068] is the time window function, It is a frequency variable, and different frequencies correspond to different oscillation periods in the wave;

[0069] S22. Extracting the main frequencies through spectrum analysis , which is the frequency with the most concentrated energy in the spectrum. First, in the spectrum Find the frequency range where the energy distribution is concentrated , and then determine the main frequency by weighted averaging:

[0070] ;

[0071] in, is the energy of the corresponding frequency. This formula calculates the energy-weighted average frequency within the frequency range, comprehensively considers the energy of all frequencies within the range, and obtains the main frequency that reflects the main oscillation characteristics of the wave.

[0072] Calculate the extracted main frequencies The reciprocal of gets the dominant period of the wave , and calculate the response time window of the unmanned ship based on the dominant cycle ,This response time window determines the shortest time for the unmanned ship to complete the adjustment before the wave impact.

[0073] S23. Use filters to further filter out noise and adjust the filter coefficients based on real-time wave data , optimize the detection effect of wave signals:

[0074] .

[0075] S3. The wave feature matrix X Input the LSTM model, and through training on historical wave data, learn the characteristics and changing trends in wave time series data, and predict the height, speed and direction of future waves. By optimizing the loss function of the model, minimize the error between the predicted value and the actual value, and continuously adjust the model parameters.

[0076] S31. The wave characteristic matrix Input LSTM model where is the time domain signal of the sensor after filtering, Represents characteristic parameters extracted from the physical model, including the main frequencies and dominant cycle Etc. Wave Characteristics Matrix It reflects the dynamic characteristics of the waves and constitutes the input layer.

[0077] S32. Hidden layer: In the LSTM model, the input features are recursively processed through multiple layers of LSTM units to capture the wave timing characteristics and calculate the hidden state matrix Obtained by recursive calculation:

[0078] ;

[0079] in, is the hidden layer weight matrix, is the weight matrix from the input layer to the hidden layer, is the input feature of the current time step, is the hidden state of the previous time step, is the bias term, is the activation function.

[0080] The bidirectional LSTM unit processes the sequence data forward and backward, and the hidden state of the final output is the concatenation of the forward and reverse LSTMs.

[0081] S33. Hidden layer output state Transformed into predicted wave state through a fully connected layer:

[0082] ;

[0083] in:

[0084] is the weight matrix from the hidden layer to the output layer, is the bias term of the output layer;

[0085] It is a prediction of the wave state at future time steps, including the height, direction and speed of the wave.

[0086] S34. Calculate predicted value The actual wave state The mean square error between:

[0087]

[0088] in:

[0089] It is the 2-norm of the difference between the true value and the predicted value. The specific calculation formula is:

[0090] ;

[0091] The real wave state Features, is the corresponding predicted value. Through the back propagation algorithm, according to the error For model parameters Perform gradient descent optimization to gradually reduce the prediction error and improve the prediction accuracy of the model.

[0092] S4. According to the predicted future wave state, the controller adjusts the heading and speed of the unmanned ship in real time. When the wave height, speed or direction exceeds the set threshold, the adjusted control command is dynamically generated to make the unmanned ship adaptively adjust and maintain stable navigation;

[0093] S41. Create / wave_sensor_data node to subscribe to the real-time sensor data topic about wave status, and preprocess it to obtain . Using the LSTM model to Features are used to predict and calculate future wave states The predicted wave state is published as a ROS message to a new topic ( / predicted_wave_state) for control adjustment of the unmanned ship.

[0094] S42. Create a control node ( / wave_adaptive_control), subscribe to the / predicted_wave_state topic, and adaptively adjust the heading and speed of the unmanned ship according to the wave prediction results.

[0095] Condition 1: When the wave height Greater than high threshold (e.g. 1.5m), the unmanned ship should reduce its speed to reduce the impact of waves on the hull; the new speed Current speed Subtract a certain percentage:

[0096] ;

[0097] in is the deceleration factor (e.g. 0.3).

[0098] Condition 2: When the wave height is below the low threshold When the wave is smaller than 0.5m, the impact of waves on navigation is small, and the speed can be increased appropriately to speed up the journey:

[0099] ;

[0100] in is the speed-up factor (e.g. 0.1).

[0101] Condition 3: When the wave direction Heading with the unmanned ship The difference is greater than the set angle threshold (e.g. 30°), the new heading Adjust to be consistent with or close to the wave direction to reduce lateral forces:

[0102] .

[0103] Condition 4: When the wave speed Exceeding speed threshold (For example, 2m / s), it means that the wave energy is high and has a greater impact on the stability of the hull. At this time, the speed should be reduced to enhance the hull control ability:

[0104] ;

[0105] in is the deceleration factor (e.g. 0.2).

[0106] S43. The / wave_adaptive_control control node generates the adjusted heading and speed control commands, which are published to the controller ( / cmd_control) in the form of ROS messages. Then they are passed to the motion controller of the unmanned ship. The unmanned ship adjusts the heading according to the received desired direction. and expected speed Instructions are given to dynamically adjust the rudder angle and propeller speed of the ship to achieve the desired heading and speed.

[0107] S5. Use a high-dimensional Kalman filter to feed back the actual state of the unmanned ship to the virtual model, optimize the control strategy based on the feedback information, ensure the synchronization of the virtual and real states, and ensure that the unmanned ship maintains stability and navigation efficiency under wave interference.

[0108] S51. When executing control instructions, the unmanned ship publishes its real-time status (including current heading, speed, attitude, etc.) in the form of ROS messages to the / ship_state topic, which constitutes the actual status of the unmanned ship. .

[0109] S52. Using the multi-dimensional real-time feedback of the high-dimensional Kalman filter, the actual state of the unmanned ship is Feedback to the virtual model to compare the virtual model state estimates and actual status , calculate the error and correct the virtual model according to the real-time error and , the control node ( / wave_adaptive_control) dynamically adjusts the control strategy.

[0110] When the speed error When the speed exceeds the expected speed by 15%, the deceleration factor Increase by 0.1-0.2, reduce the speed of the unmanned ship to ensure stable navigation. When the speed error is less than 5% of the expected speed, reduce the deceleration factor or increase the acceleration factor , in order to increase the speed of unmanned ships and enhance navigation efficiency.

[0111] When the heading error When the heading error exceeds 15°, the sensitivity of the heading adjustment is increased, and the adjustment step is increased by 10-30%, so as to accelerate the steering of the unmanned boat to sail against the waves. When the heading error is less than 5°, the sensitivity is reduced to prevent over-adjustment and ensure the smooth sailing of the unmanned boat.

[0112] The unmanned ship adaptive wave disturbance control system based on digital twin includes:

[0113] Wave information collection module: collects data such as wave height, speed and direction in real time, and provides basic information about the sea conditions around the unmanned ship;

[0114] Data preprocessing and feature extraction module: Filter and denoise the collected wave data, extract the main frequencies and characteristic parameters, and construct the wave feature matrix for subsequent model prediction;

[0115] Wave state prediction module: Use the LSTM model to train and predict the wave feature matrix, generate prediction results of future wave states, and support real-time decision-making of unmanned ships;

[0116] Unmanned ship control decision module: outputs unmanned ship control instructions, formulates the unmanned ship's heading and speed adjustment strategy according to the wave prediction results, and ensures that the unmanned ship can adapt to the wave environment and maintain stable navigation;

[0117] Virtual-reality feedback and optimization module: The actual navigation status of the unmanned ship is fed back to the digital twin model, and the virtual-reality state synchronization is achieved through a high-dimensional Kalman filter. According to the optimized control strategy, the unmanned ship's adaptive optimization control is achieved.

[0118] Although the embodiments of the present application have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present application, and that the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. An adaptive wave disturbance control method for unmanned ships based on digital twins, characterized in that: The following steps are involved: S1. Collect wave signals through wave radar sensors to obtain time domain information of wave height, wavelength, wave speed and direction; S2. Preprocess the collected wave signals, remove noise and extract the main frequency and dominant period features to form a wave feature matrix X; S21. Perform frequency domain analysis on the wave signal S(t) using generalized Fourier transform and convert it into a wave spectrum S(f) in the frequency domain; S22. Extract the main frequency f through spectrum analysis dominat , which is the frequency with the most concentrated energy in the spectrum; first, find the frequency range in which the energy distribution is concentrated in the spectrum |S(f)| lower ,f upper ], and then determine the main frequency by weighted averaging: Among them, |S(f)| is the energy of the corresponding frequency, f lower is the minimum value of the frequency variable f, f upper is the maximum value of the frequency variable f; Calculate the extracted main frequency f dominant The reciprocal of the wave gives the dominant period T wave , and calculate the response time window T of the unmanned ship according to the dominant cycle response ,This response time window determines the shortest time for the unmanned ship to complete the adjustment before the wave impact; S3. Input the wave feature matrix X into the LSTM model, use historical wave data for training, learn the characteristics and change trends in the wave time series data, and predict the height, speed and direction of future waves; minimize the error between the predicted value and the actual value by optimizing the loss function of the model, and continuously adjust the model parameters; S4. According to the predicted future wave state, the controller obtains the wave prediction results and adjusts the heading and speed of the unmanned ship in real time; when the wave height, speed or direction exceeds the set threshold, the adjusted control command is dynamically generated to make the unmanned ship adaptively adjust; S5. Use a high-dimensional Kalman filter to feed back the actual state of the unmanned ship to the virtual model, optimize the control strategy based on the feedback information, ensure the synchronization of the virtual and real states, and ensure that the unmanned ship maintains stability and navigation efficiency under wave interference.

2. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: The training process of the LSTM model includes: S3.1 The wave characteristic matrix X = [S filtered (t),P] input LSTM model, where S filtered (t) is the filtered sensor time domain signal, P represents the characteristic parameters extracted from the physical model, including the main frequency and dominant period, describing the dynamic characteristics of the wave; S3.2 processes the input feature matrix X through multiple layers of LSTM units and calculates the hidden state h t , recursively capture the wave timing characteristics, and use bidirectional LSTM to process the sequence data forward and backward to obtain the complete hidden state; S3.3 The hidden state h t Input to the fully connected layer to generate predictions for future wave conditions including the height, speed and direction of the waves; S3.4 Calculate the predicted value The error between the actual wave state Y and the model is optimized by minimizing the loss function; S3.5 uses the back propagation algorithm to adjust the model parameters and perform gradient descent optimization on the weight matrix.

3. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: In step S4, S41. Use LSTM model to calculate future wave state The controller issues wave prediction instructions to each control node in the form of ROS messages for control adjustment of the unmanned ship; S42. The control node obtains the wave prediction instruction and adaptively adjusts the heading and speed of the unmanned ship according to the wave prediction result; S43. The unmanned ship receives the heading θ new and speed V new Instructions to dynamically adjust the rudder angle and propeller speed of the ship.

4. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: In step S4, the unmanned ship is adaptively adjusted as follows: Condition 1: When the wave height H wave Greater than the high threshold H threshold_high When the unmanned ship is at a lower speed, it should reduce the impact of waves on the hull. The new speed V new The calculation is as follows: In new =V current ×(1-k1) Where k1 is the deceleration coefficient; Condition 2: When the wave height is lower than the low threshold H threshold_low When the new speed V new The calculation is as follows: V new =V current ×(1+k2) Where k2 is the growth rate coefficient; Condition 3: When the wave direction is D wave and the unmanned ship heading θ ship The phase difference is greater than the set angle threshold θ threshold , the new heading θ new Adjust to be consistent with or close to the wave direction, the new heading θ new The calculation is as follows: i new =D wave Condition 4: When the wave speed C wave Exceeding speed threshold C threshold When the wave energy is high, it means that the wave energy is high and has a greater impact on the stability of the ship. new The calculation is as follows: V new =V current ×(1-k3) Where k3 is the deceleration factor.

5. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: In step S4, when the unmanned ship executes the control command, it publishes its real-time status to the control node in the form of ROS message, which constitutes the actual status of the unmanned ship. physical .

6. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 5 is characterized in that: In step S5, the actual state D of the unmanned ship is converted into physical Feedback to the virtual model to compare the virtual model state estimates and the actual state z t , calculate the error and correct the virtual model to ensure the consistency between the virtual model and the physical unmanned ship.

7. The method for adaptive wave disturbance control of unmanned ships based on digital twins according to claim 6 is characterized in that: in, According to the real-time error V =V new -V actual and e θ =θ new -θ actual , the control node dynamically adjusts the control strategy; V actual is the actual speed, θ actual is the actual heading; When the speed error e V Exceeding speed V new When the speed error is less than 15%, increase the deceleration coefficients k1 and k3; when the speed error is less than the speed V new When the speed reaches 5%, reduce the deceleration coefficient k1, k3 or increase the speed increase coefficient k2. When the heading error e θ When the angle exceeds 15°, increase the step size.

Images