Adaptive Wave Interference Control Method and System for Unmanned Vessels Based on Digital Twin

Abstract

This application belongs to the field of unmanned vessel control technology, specifically to a method and system for adaptive wave interference control of unmanned vessels based on digital twins, including a data acquisition module, a data processing module, a decision module, and an output module. Its advantage lies in the use of an adaptive multidimensional regression strategy to evaluate the differences between the virtual model and the physical model, and to dynamically adjust the system based on the prediction and feedback results. While ensuring the navigation safety and mission execution of the unmanned vessel, it improves the efficiency of mission execution and the robustness of the system, and enhances the adaptability and continuity of digital twin technology in the complex mission environment of unmanned vessels.

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 and system based on digital twins. Background Art

[0002] Unmanned vessels are surface robots that can navigate waters according to pre-set missions without remote control, relying on precise satellite positioning and self-sensing capabilities. These "surface robots" integrate multiple technologies, including marine, communications, automatic control, remote monitoring, and networked systems, enabling autonomous navigation, intelligent obstacle avoidance, long-distance communication, real-time video transmission, and networked monitoring. As a new technological tool, unmanned vessels have been widely used in fields such as oceanographic surveying and maritime defense.

[0003] Unmanned vessels often encounter high sea conditions caused by wind and waves during their operation. In these conditions, unmanned vessels are prone to capsizing, stalling, or even running away. Therefore, the control methods used for unmanned vessels in high sea conditions are directly related to their operational safety and mission suitability. Currently, unmanned vessel control methods can ensure stable operation in low sea conditions. However, when encountering high sea conditions, existing unmanned vessel control methods are unable to adaptively control the unmanned vessel based on the sea conditions, and therefore cannot guarantee stable operation in these conditions. Summary of the Invention

[0004] Based on the above problems, this application provides a method and system for adaptive wave interference control of unmanned ships based on digital twins, and its technical solution is as follows:

[0005] An adaptive wave interference control method for an unmanned ship based on digital twins includes the following steps:

[0006] S1. Obtain wave information and perform wave frequency analysis;

[0007] S2. Calculate response time;

[0008] S3. Control the attitude and speed of the unmanned vessel;

[0009] S4. Adaptive posture and speed optimization of unmanned vessels using deep reinforcement learning.

[0010] Preferably, in step S1,

[0011] S11. Use sensors to collect wave signals in real time , including wave height ,wavelength , wave speed and wave direction Time domain information, this time domain signal Reflects the changes in the time dimension of the wave:

[0012] ;

[0013] S12. Using generalized Fourier transform to analyze wave signals Perform frequency domain analysis and convert it into wave spectrum in the frequency domain :

[0014] ;

[0015] is the time window function, is a frequency variable, different frequency components Corresponding to different oscillation periods in the waves;

[0016] S13. Extracting the main frequencies through frequency spectrum analysis , that is, the frequency with the most concentrated energy in the spectrum. The main frequency is determined by finding the frequency peak with the largest amplitude in the spectrum:

[0017] ;

[0018] In the wave spectrum In the process, by scanning the frequency range, find the area with the largest energy, that is, the location of the main frequency;

[0019] S14. 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:

[0020] .

[0021] Preferably, in step S1, in the case of multi-frequency waves, the frequency bandwidth is calculated , which represents the frequency range between frequency components close to the main frequency, and is used to capture the energy concentration within a wider frequency band:

[0022] ;

[0023] and They are the upper and lower frequency limits, respectively, indicating the frequency range with greater fluctuation intensity in the wave spectrum.

[0024] Preferably, S21. According to the frequency of extraction Calculate the main angular frequencies of the waves :

[0025] ;

[0026] Calculate the dominant period of the wave :

[0027] ;

[0028] S22. Using the calculated main frequency and dominant cycles , calculate the response time window of the unmanned ship ,This window determines the shortest time the unmanned vessel needs to complete the adjustment before the wave hits;

[0029] ;

[0030] in, is the safety factor.

[0031] Preferably, in step S3, the attitude control method is as follows: by using the optimal control theory, the optimal attitude adjustment strategy of the unmanned ship is calculated according to the wave interference to ensure that the parameters such as the pitch angle and the roll angle of the hull can minimize the influence of the waves. The control equation is:

[0032] ;

[0033] in is the angular momentum of the unmanned ship, is the wave torque, Control torque for thrusters;

[0034] By solving the Hamiltonian equation, the optimal control strategy of the unmanned ship is obtained:

[0035] ;

[0036] in, To solve the matrix, is the attitude control input, is the control input weight matrix, is the control gain matrix, is the attitude angle of the unmanned vessel;

[0037] The speed control method is as follows:

[0038] Calculate the optimal output power of the thruster , so that the speed of the unmanned ship can smoothly cope with the impact of waves:

[0039] ;

[0040] in, For the quality of unmanned ships, is the wave impact force, The speed of the unmanned vessel.

[0041] Preferably, in step S4, the attitude and speed of the unmanned boat are adaptively adjusted through a multi-objective optimization strategy, and the reward function Combined with wave impact , posture adjustment and energy consumption To optimize:

[0042] ;

[0043] 、 、 Respectively represent and The attitude adjustment includes the attitude angle adjustment (including pitch angle, roll angle and yaw angle) and spatial position adjustment (including displacement and relative position) of the unmanned ship.

[0044] Preferably, an emergency avoidance and recovery strategy of a multi-objective evolutionary algorithm is also included, and the steps are as follows:

[0045] S5.1 When the wave impact Exceeding safety threshold When , a multi-objective evolutionary algorithm is used for emergency avoidance; the objective function includes wave intensity, avoidance path cost and task execution cost:

[0046] ;

[0047] and is the weight coefficient in multi-objective optimization;

[0048] S5.2 is solved by a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss.

[0049] Preferably, the actual state of the unmanned ship is reflected by the multi-dimensional real-time feedback of the digital twin and the high-dimensional Kalman filter. Feedback is fed back to the virtual model, and the virtual model is adaptively corrected through a high-dimensional Kalman filter:

[0050] ;

[0051] in, is the Kalman gain, is the virtual model state estimation, is the control parameter of the system, The real-time measurement value of the physical world, that is, the actual state of the unmanned ship ; is the state transition matrix, is the measurement matrix, Control input matrix.

[0052] An unmanned vessel adaptive wave interference control system based on digital twin is used to execute an unmanned vessel adaptive wave interference control method based on digital twin, comprising a data acquisition module, a data processing module, a decision module and an output module;

[0053] Data acquisition module: acquires physical information data of the electromechanical product under study through sensors, monitoring equipment or other data sources, including sensor data, operation records, and environmental parameter data;

[0054] Data processing module: Exchanges the collected multi-source data into the digital space, performs data preprocessing and analysis, and obtains key modeling parameters for working conditions and constraints;

[0055] Decision-making module: Based on the provided measured data sets of physical space and simulated data sets of digital space, a deep learning model is used for training. An emergency avoidance and recovery strategy based on the target optimization algorithm is constructed and solved using a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss.

[0056] Output module: Utilizes the multi-dimensional real-time feedback of digital twins and high-dimensional Kalman filters to visualize the results.

[0057] Compared with the prior art, this application has the following beneficial effects:

[0058] 1. This application utilizes digital twin technology to achieve real-time synchronization between the physical system and the virtual model, and uses dynamic feedback from the virtual model to modify the state of the physical unmanned vessel. Compared to existing control methods that rely solely on physical sensors, this invention can more accurately predict the impact of wave disturbances on the unmanned vessel, enabling proactive adjustment measures, thereby improving the system's response speed and accuracy.

[0059] 2. By combining the generalized Fourier transform and adaptive filtering, this application can more accurately capture the primary frequency components in complex wave environments, improving the accuracy of wave interference prediction. Compared with traditional time-domain analysis, this invention can more effectively cope with multi-frequency wave interference, especially in sudden and non-stationary wave environments.

[0060] 3. This invention utilizes a multi-objective evolutionary algorithm to simultaneously optimize the UAV's avoidance path and mission execution efficiency when faced with wave disturbances. Unlike existing single-objective control methods, this method dynamically balances wave disturbance intensity, avoidance path safety, and mission completion, enabling the UAV to maintain efficient operation in complex sea conditions.

[0061] 4. By leveraging deep reinforcement learning technology, this invention can adaptively optimize the unmanned vessel's attitude and speed adjustment strategies based on real-time wave conditions. Compared to traditional pre-set control schemes, deep reinforcement learning can self-optimize according to changing sea conditions, significantly improving the stability and energy efficiency of the unmanned vessel. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is the flow chart of this application. DETAILED DESCRIPTION

[0063] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the embodiments of the present invention and the accompanying drawings. Obviously, the embodiments described 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 making creative efforts are within the scope of protection of the present invention.

[0064] An adaptive wave interference control method for an unmanned ship based on digital twins includes the following steps:

[0065] S1. Obtain wave information and perform wave frequency analysis;

[0066] Step S1 includes:

[0067] S11. Use sensors to collect wave signals in real time , including wave height ,wavelength , wave speed and wave direction Time domain information, this time domain signal Reflects the changes in the time dimension of the wave:

[0068] ;

[0069] S12. Using generalized Fourier transform to analyze wave signals Perform frequency domain analysis and convert it into wave spectrum in the frequency domain :

[0070] ;

[0071] is the time window function, is a frequency variable, different frequency components Corresponding to different oscillation periods in the waves;

[0072] S13. Extracting the main frequencies through frequency spectrum analysis , that is, the frequency with the most concentrated energy in the spectrum. The main frequency is determined by finding the frequency peak with the largest amplitude in the spectrum:

[0073] ;

[0074] In the wave spectrum In the process, by scanning the frequency range, find the area with the largest energy, that is, the location of the main frequency;

[0075] S14. 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:

[0076] .

[0077] Calculate the frequency bandwidth in the case of multi-frequency waves , which represents the frequency range between frequency components close to the main frequency, and is used to capture the energy concentration within a wider frequency band:

[0078] ;

[0079] and They are the upper and lower frequency limits, respectively, indicating the frequency range with greater fluctuation intensity in the wave spectrum.

[0080] S2. Calculate response time;

[0081] S21. According to the frequency of extraction Calculate the main angular frequencies of the waves :

[0082] ;

[0083] Calculate the dominant period of the wave :

[0084] ;

[0085] S22. Using the calculated main frequency and dominant cycles , calculate the response time window of the unmanned ship ,This window determines the shortest time the unmanned vessel needs to complete the adjustment before the wave hits;

[0086] ;

[0087] in, is the safety factor.

[0088] S3. Control the attitude and speed of the unmanned vessel;

[0089] In step S3, the attitude control method is as follows: through the optimal control theory, the optimal attitude adjustment strategy of the unmanned ship is calculated according to the wave interference to ensure that the parameters such as the hull's pitch angle and roll angle can minimize the impact of the waves. The control equation is:

[0090] ;

[0091] in is the angular momentum of the unmanned ship, is the wave torque, Control torque for thrusters;

[0092] By solving the Hamiltonian equation, the optimal control strategy of the unmanned ship is obtained:

[0093] ;

[0094] in, To solve the matrix, is the attitude control input, is the control input weight matrix, is the control gain matrix, is the attitude angle of the unmanned vessel;

[0095] The speed control method is as follows:

[0096] Calculate the optimal output power of the thruster , so that the speed of the unmanned ship can smoothly cope with the impact of waves:

[0097] ;

[0098] in, For the quality of unmanned ships, is the wave impact force, The speed of the unmanned vessel.

[0099] S4. Adaptive posture and speed optimization of unmanned vessels using deep reinforcement learning.

[0100] The attitude and speed of the unmanned boat are adaptively adjusted through a multi-objective optimization strategy, and the reward function Combined with wave impact , posture adjustment and energy consumption To optimize:

[0101] ;

[0102] 、 、 Respectively represent and The attitude adjustment includes the attitude angle adjustment (including pitch angle, roll angle and yaw angle) and spatial position adjustment (including displacement and relative position) of the unmanned ship.

[0103] S5. Emergency avoidance and recovery strategy of multi-objective evolutionary algorithm, the steps are as follows:

[0104] S5.1 When the wave impact Exceeding safety threshold When , a multi-objective evolutionary algorithm is used for emergency avoidance; the objective function includes wave intensity, avoidance path cost and task execution cost:

[0105] ;

[0106] and is the weight coefficient in multi-objective optimization.

[0107] S5.2 is solved by a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss.

[0108] The actual state of the unmanned ship is reflected by the multi-dimensional real-time feedback of digital twin and high-dimensional Kalman filter. Feedback is fed back to the virtual model, and the virtual model is adaptively corrected through the high-dimensional Kalman filter:

[0109] ;

[0110] in, is the Kalman gain, is the virtual model state estimation, is the control parameter of the system, The real-time measurement value of the physical world, that is, the actual state of the unmanned ship ; is the state transition matrix, is the measurement matrix, Control input matrix.

[0111] An unmanned vessel adaptive wave interference control system based on digital twin, used for an unmanned vessel adaptive wave interference control method based on digital twin, comprising a data acquisition module, a data processing module, a decision module and an output module;

[0112] Data acquisition module: acquires physical information data of the electromechanical product under study through sensors, monitoring equipment or other data sources, including sensor data, operation records, and environmental parameter data;

[0113] Data processing module: Exchanges the collected multi-source data into the digital space, performs data preprocessing and analysis, and obtains key modeling parameters for working conditions and constraints;

[0114] Decision-making module: Based on the provided measured data sets of physical space and simulated data sets of digital space, a deep learning model is used for training. An emergency avoidance and recovery strategy based on the target optimization algorithm is constructed and solved using a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss.

[0115] Output module: Utilizes the multi-dimensional real-time feedback of digital twins and high-dimensional Kalman filters to visualize the results.

[0116] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for adaptive wave interference control of unmanned ships based on digital twins, characterized in that: The following steps are involved: S1. Obtain wave information and perform wave frequency analysis; S2. Calculate response time; S3. Control the attitude and speed of the unmanned vessel; The attitude control method is as follows: Through the optimal control theory, the optimal attitude adjustment strategy of the unmanned ship is calculated according to the wave interference to ensure that the hull's pitch angle and roll angle parameters can minimize the impact of waves. The control equation is: Where L is the angular momentum of the unmanned ship, τ wave is the wave torque, T thruster Control torque for thrusters; By solving the Hamiltonian equation, the optimal control strategy of the unmanned ship is obtained: u opt =-R -1 G T Pθ; Among them, P is the solution matrix, u opt is the attitude control input, R is the control input weight matrix, G is the control gain matrix, and θ is the attitude angle of the unmanned ship; The speed control method is as follows: Calculate the optimal output power F of the thruster thruster,opt , so that the speed of the unmanned ship can smoothly cope with the impact of waves: Among them, m is the mass of the unmanned ship, F wave is the wave impact force, V is the speed of the unmanned ship; S4. Adaptive posture and speed optimization of unmanned vessels using deep reinforcement learning; The attitude and speed of the unmanned boat are adaptively adjusted through a multi-objective optimization strategy. The reward function R(t) is combined with the wave impact force F wave , attitude adjustment ΔQ(t) and energy consumption E(t) are optimized: R(t)=-α1F wave (t)-α2ΔQ(t)-α3E(t); α1, α2, and α3 represent F wave The weight coefficients of (t), ΔQ(t) and E(t).

2. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: In step S1, S11. Use sensors to collect wave signals S(t) in real time, including wave height H t , wavelength L t , wave speed C t and wave direction D t The time domain information of the wave is represented by the time domain signal H(t). The time domain signal H(t) reflects the change of the wave in the time dimension: S(t)={H t (t),L t (t),C t (t),D t (t)}; S12. Use generalized Fourier transform to perform frequency domain analysis on the wave signal S(t) and convert it into a wave spectrum S(f) in the frequency domain: g(t) is the time window function, f is the frequency variable, and different frequency components f correspond to different oscillation periods in the wave; S13. Extract the main frequency f through frequency spectrum analysis dominant , that is, the frequency with the most concentrated energy in the spectrum. The main frequency is determined by finding the frequency peak with the largest amplitude in the spectrum: In the wave spectrum |S(f)|, by scanning the frequency range, find the area with the largest energy, that is, the location of the main frequency; S14. Use a filter to further filter out noise and adjust the filter coefficient W(t) according to the real-time wave data to optimize the detection effect of the wave signal: S filtered (t)=W(t)*S(t)。 3. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 2 is characterized in that: In step S1, in the case of multi-frequency waves, the frequency bandwidth Δf is calculated, which represents the frequency range between frequency components close to the main frequency: Δf=f upper -f lower ; f upper and f lower are the upper and lower frequency limits respectively.

4. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: S21. According to the extracted frequency f dominant Calculate the main angular frequency ω of the wave dominat : oh dominat =2πf dominant ; Calculate the dominant period T of the wave wave : S22. Using the calculated main frequency ω dominat and the dominant period T wave , calculate the response time window T of the unmanned ship response ,This window determines the shortest time the unmanned ship needs to complete the adjustment before the wave hits; Where η is the safety factor.

5. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: It also includes the emergency avoidance and recovery strategy of the multi-objective evolutionary algorithm. The steps are as follows: S5.1 When the wave impact force F wave (t+k) exceeds the safety threshold ∈ safe When , a multi-objective evolutionary algorithm is used for emergency avoidance; the objective function includes wave intensity, avoidance path cost and task execution cost: Y=min(C path (t)+β1F wave (t)+β2C task (t)); β1 and β2 are weight coefficients in multi-objective optimization; S5.2 is solved by a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss.

6. The method for adaptive wave interference control of unmanned ships based on digital twins according to claim 1 is characterized in that: The actual state of the unmanned ship is D physical Feedback is fed back to the virtual model, and the virtual model is adaptively corrected through the high-dimensional Kalman filter: Among them, K t is the Kalman gain, is the virtual model state estimation, u t is the control parameter of the system, z t is the real-time measurement value of the physical world, i.e. the actual state of the unmanned ship D physical ; is the state transition matrix, is the measurement matrix, Control input matrix.

7. An unmanned vessel adaptive wave interference control system based on digital twin, used to implement the unmanned vessel adaptive wave interference control method based on digital twin according to any one of claims 1 to 6, characterized in that: It includes data acquisition module, data processing module, decision module and output module; Data acquisition module: acquires physical information data of the electromechanical product under study through sensors, monitoring equipment or other data sources, including sensor data, operation records, and environmental parameter data; Data processing module: Exchanges the collected multi-source data into the digital space, performs data preprocessing and analysis, and obtains key modeling parameters for working conditions and constraints; Decision-making module: Based on the provided measured data sets of physical space and simulated data sets of digital space, a deep learning model is used for training. An emergency avoidance and recovery strategy based on the target optimization algorithm is constructed and solved using a non-dominated sorting genetic algorithm or a fast non-dominated sorting genetic algorithm to select the optimal avoidance path to maximize safety and minimize mission loss. Output module: Utilizes the multi-dimensional real-time feedback of digital twins and high-dimensional Kalman filters to visualize the results.

Images