Iterative learning unmanned surface vessel course control method and system based on data encryption

Abstract

The invention provides an iterative learning unmanned surface vessel course control method and system based on data encryption, and relates to the technical field of unmanned surface vessel course control, and the method comprises the steps: redefining the output of an unmanned surface vessel course control system; deducing an output increment through output redefinition, determining a pseudo partial derivative, and establishing a compact-form dynamic linearization model; establishing an encoding-decoding mechanism; updating control input through iterative learning on the basis of the compact-form dynamic linearization model and encoding and decoding data; the rudder angle of the unmanned surface vessel is adjusted according to the control input, and course control is achieved. According to the method, the problems of high model dependence, quasi-linear hypothesis conflict, insufficient data utilization, potential safety hazards and the like in the prior art are effectively solved, the accuracy, efficiency and safety of course control of the unmanned surface vessel are remarkably improved, and an innovative solution is provided for unmanned system control in a dynamic uncertain environment.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned surface vessel heading control, and in particular to a data-based encrypted iterative learning heading control method and system for an unmanned surface vessel. Background Art

[0002] Unmanned surface vehicles (USVs) are widely used in marine environmental monitoring, underwater topography mapping, and military reconnaissance due to their high maneuverability, strong mission adaptability, and low-risk operation capabilities. However, existing USV heading control technology still has significant shortcomings, specifically in the following three aspects:

[0003] 1. Limitations of relying on precise mathematical models: Existing mainstream methods (such as sliding mode control and model predictive control) rely on precise mathematical models of USVs. However, in complex and dynamic ocean environments (such as wind, current, and wave disturbances), accurate modeling is difficult, resulting in degraded control performance and insufficient robustness.

[0004] 2. Inapplicability of the quasi-linearization assumption: Model-free adaptive control (MFAC) methods usually assume that the system has quasi-linear characteristics (that is, an increase in control input does not lead to a decrease in output). However, in USV heading control, the rudder angle input and heading angle output often show nonlinear or even reverse changes due to inertia or environmental interference, making existing MFAC methods difficult to directly apply.

[0005] 3. Application gaps in iterative learning control: Although USVs often perform repetitive tasks and accumulate a large amount of historical data, existing methods (such as PID control and reinforcement learning) do not fully utilize this data. Although model-free adaptive iterative learning control (MFAILC) has been applied in trains and multi-agent systems, its research on USV heading control is still in its early stages, and its convergence speed and stability need to be improved.

[0006] 4. Inadequate information security and communication efficiency: Lightweight USV communication equipment has strict bandwidth requirements. Existing encryption mechanisms (such as static encoding and decoding schemes) are mostly designed for known dynamic systems and are difficult to adapt to the dynamic encryption requirements of USVs in uncertain environments. Existing methods do not combine MFAILC with encryption technology, resulting in low data transmission efficiency and vulnerability to cyber attacks.

[0007] In summary, there is an urgent need for a USV heading control method with high precision, high efficiency and high safety. Summary of the Invention

[0008] To this end, an embodiment of the present invention provides an iterative learning unmanned surface vehicle heading control method and system based on data encryption, which is used to solve the problems of strong model dependence, quasi-linear assumption conflict, insufficient data utilization and security risks in the existing technology.

[0009] In order to solve the above problems, an embodiment of the present invention provides an iterative learning unmanned surface vehicle heading control method based on data encryption, the method comprising:

[0010] S1: Redefine the output of the unmanned surface vehicle heading control system as:

[0011]

[0012] in, For the moment No. The output of the iteration, For the moment No. The heading angle of the iteration, is the angular velocity at the corresponding moment, K1 is the angular velocity gain;

[0013] S2: Derivation of output increments through the output redefinition Determine pseudopartial derivatives When satisfied When , a compact form dynamic linearization model is established:

[0014]

[0015] Where T is the system time constant, T s is the sampling period, K is the control gain, u max is the maximum input amplitude, For control input increment;

[0016] S3: Establish an encoding-decoding mechanism, including:

[0017] The logarithmic quantizer Q(·) is used to encrypt the redefined output and recursively generate the encoded cumulative value. The formula is:

[0018]

[0019] Where E represents the encoder, Υ(0,ζ) is the cumulative value of the code at the starting point of the iteration, is the accumulated value of the encoder output, is a dynamic scaling function, is the post-quantization increment;

[0020] The original data is restored based on the coded cumulative value. The formula is:

[0021]

[0022] in, To estimate the output, For the moment The estimated output of the ζth iteration;

[0023] S4: Based on the compact form dynamic linearized model and the encoded and decoded data, the control input is updated through iterative learning:

[0024]

[0025] in, is the target heading angle, ρ is the step size factor, λ is the weight coefficient, is the estimated value of the pseudo partial derivative at the current moment and iteration, Estimated output for the previous iteration;

[0026] S5: According to the control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control.

[0027] Preferably, the pseudo partial derivative estimate The update formula is:

[0028]

[0029] in, is the estimated value of the pseudo partial derivative of the previous iteration, η is the learning rate, and μ is the regularization parameter.

[0030] Preferably, the quantizer Q(·) is defined as:

[0031] Q={±x i :x i =θ i x0,i=0,±1,±2,…}∪{0},0<θ<1,x0>0;

[0032] Among them, x i is the quantized value, θ is the quantization base, and x0 is the initial quantization value.

[0033] Preferably, the dynamic scaling function satisfy And dynamically adjust according to the time step or number of iterations, including but not limited to sine scaling function

[0034] Preferably, the parameters satisfy ρ∈(0,1], η∈(0,1], μ>0, λ>λ min , where λ min The minimum weight coefficient is used, and the parameter combination is optimized through sensitivity analysis to ensure that the tracking error is bounded.

[0035] Preferably, the encoder outputs a quantized increment With the original heading angle Nonlinear correlation and dynamic scaling function The data compression and encryption are realized by the logarithmic quantizer Q(·), so that the attacker cannot infer the true heading state by intercepting the data.

[0036] An embodiment of the present invention further provides an iterative learning unmanned surface vehicle heading control system based on data encryption, which is used to implement the above-mentioned iterative learning unmanned surface vehicle heading control method based on data encryption, specifically comprising:

[0037] Sensors for collecting real-time heading angles of unmanned surface vessels and angular velocity

[0038] Redefine the formula Generate Redefinition Output

[0039] Encoder, used to encrypt the redefined output and generate the encoded cumulative value

[0040] Decoder, used to decode the encrypted data and recover the estimated output

[0041] Pseudo partial derivative updater, used to update the pseudo partial derivative estimate based on historical input and output data

[0042] Controller, used to adjust the target heading angle Estimated output and pseudo-partial derivative estimates Generate control input

[0043] Actuator, used to control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control.

[0044] Preferably, the logarithmic quantizer Q(·) and the dynamic scaling function in the encoder Supports dynamic parameter adjustment to adapt to the communication efficiency and security requirements in different marine environments.

[0045] Preferably, the controller has a built-in parameter adaptive adjustment mechanism to optimize the step size factor ρ and the weight coefficient λ in real time to ensure the robustness of the system in complex environments.

[0046] An embodiment of the present invention also provides a computer storage medium, which stores a computer software product. The computer software product includes several instructions for enabling a computer device to execute the above-mentioned iterative learning unmanned surface vehicle heading control method based on data encryption.

[0047] It can be seen from the above technical solutions that the present invention has the following beneficial effects:

[0048] (1) This paper redefines the output as a weighted combination of heading angle and angular velocity, constructs a compact form dynamic linearization (CFDL) model, introduces angular velocity to reflect the immediate dynamic effect of the input, and ensures that the pseudo partial derivative Ψ>0 always holds, completely satisfying the premise of MFAC theory. This eliminates the impact of nonlinear interference on control performance, makes the system input-output relationship more sensitive and stable, and significantly improves the heading tracking accuracy and convergence speed of the USV in complex environments such as wind, waves, and flow fields.

[0049] (2) The present invention dynamically updates the control input through historical iterative data and gradually optimizes the control strategy. It is particularly suitable for tasks such as ocean monitoring and reconnaissance with repeated routes. A coding-decoding scheme based on logarithmic quantization is designed to perform nonlinear compression on the redefined output (the quantized value is distributed in a geometric series), and the encryption parameters are obfuscated by a dynamic scaling function. Historical data is used to improve tracking accuracy and avoid the cumulative error caused by model uncertainty in traditional methods. The coded output is nonlinearly related to the true heading angle. Even if an attacker intercepts the data, he cannot reversely restore the true state due to the lack of dynamic parameters.

[0050] (3) The logarithmic quantizer converts large-scale dynamic signals into small-bit-width data, which reduces the communication load compared to traditional uniform quantization. The controller parameters (step factor, learning rate, etc.) are dynamically adjusted according to the real-time status, and sensitivity analysis is combined to provide parameter optimization guidance. It is suitable for USV low-bandwidth communication scenarios, reduces data transmission delays, and ensures real-time control. The adaptive adjustment of parameters enables the system to remain stable under different sea conditions (such as changes in wind and wave levels), and enhances its anti-interference ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the implementation cases of the present invention or the technical solutions in the prior art, the following is a brief description of the drawings required for use in the embodiments. By referring to the drawings, the features and advantages of the present invention will be more clearly understood. The drawings are schematic and should not be understood as limiting the present invention in any way. Those skilled in the art can derive other drawings based on these drawings without inventive effort. Among them:

[0052] Figure 1 A flow chart of an iterative learning unmanned surface vehicle heading control method based on data encryption provided by the present invention;

[0053] Figure 2 It is an experimental effect diagram of the method of the present invention in the embodiment;

[0054] Figure 3 This is a diagram showing the effect of sensitivity analysis of parameters in the embodiment;

[0055] Figure 4 This is a structural diagram of an iterative learning unmanned surface vessel heading control system based on data encryption provided by the present invention. DETAILED DESCRIPTION

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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 making creative efforts shall fall within the scope of protection of the present invention.

[0057] Example 1

[0058] In order to solve the problems of strong model dependence, quasi-linear assumption conflict, insufficient data utilization and security risks in existing technologies, such as Figure 1 As shown, the present invention proposes an iterative learning unmanned surface vehicle heading control method based on data encryption, the method comprising:

[0059] S1: Redefine the output of the unmanned surface vehicle heading control system;

[0060] S2: Derivation of output increments by output redefinition Determine pseudopartial derivatives When satisfied When , a compact form dynamic linearization model is established;

[0061] S3: Establish an encoding-decoding mechanism, including:

[0062] The logarithmic quantizer Q(·) is used to encrypt the redefined output, recursively generate the encoded cumulative value, and restore the original data based on the encoded cumulative value;

[0063] S4: Updates the control input through iterative learning based on a compact dynamic linearized model and encoded and decoded data;

[0064] S5: Adjust the rudder angle of the unmanned surface vehicle according to the control input to achieve heading control.

[0065] As can be seen from the above technical solution, the present invention proposes an iterative learning heading control method for unmanned surface vehicles based on data encryption. By redefining the heading control system output as a weighted combination of heading angle and angular velocity, a compact form dynamic linearization (CFDL) model is constructed to solve the problem of quasi-linearization assumption failure in model-free adaptive control. A coding-decoding mechanism based on logarithmic quantization is designed, and dynamic scaling functions and logarithmic quantizers are used to compress data bits and enhance transmission security. Combining iterative learning control (ILC) with adaptive parameter adjustment, an encrypted iterative learning controller is designed. The control input is optimized based on historical iterative data and the pseudo-partial derivative estimates are adaptively adjusted. Finally, the rudder angle of the unmanned surface vehicle is adjusted according to the control input to achieve heading control. The advantages of this scheme are: eliminating the quasi-linearization constraints of traditional control methods and improving control robustness in complex environments; combining logarithmic quantization with dynamic encryption to reduce communication load while ensuring data privacy; and utilizing iterative learning to fully exploit the value of repeated task data, significantly improving heading tracking accuracy and convergence speed, providing an innovative solution for the efficient and safe operation of unmanned surface vehicles in dynamic and uncertain environments.

[0066] The present invention uses the "Dolphin I" small unmanned surface vehicle for the experiment. It is 2.0 meters long, 1.0 meters wide, has a displacement of 55.0 kilograms, and a maximum speed of about 1.5 meters per second. The dynamic parameters set in the experiment are K = 186, T = 1.068, and K1 = 10. The controller parameters are set to η=1, μ=0.3, ρ=1, λ=5, ε=10 -4 , and θ=0.8. The time interval is [0,T d ], where T d =80. The initial condition is set as and In addition, the control inputs and outputs are initialized to zero at the first iteration, i.e.

[0067] In step S1, the heading control system output is redefined:

[0068]

[0069] in, For the moment No. The output of the iteration, For the moment No. The heading angle of the iteration, is the angular velocity at the corresponding moment, K1>K min is the angular velocity gain.

[0070] Obtain the real-time heading angle of the "Dolphin I" unmanned surface vehicle through sensors and angular velocity Combined with the setting K1=10, the redefined output is calculated This redefinition makes the system input-output relationship more consistent with control requirements and lays the foundation for subsequent control processes.

[0071] In step S2, the output increment is derived based on the above output redefinition Then determine the pseudo partial derivative (PPD) Since K1 is set to 10, (Here we assume that T s ,K,u max When the parameters meet the conditions in this scenario, The following assumptions are met to establish a compact form dynamic linearization (CFDL) model:

[0072]

[0073] Where T is the system time constant, T s is the sampling period, K is the control gain, u max is the maximum input amplitude, For control input increment;

[0074] The present invention alleviates the quasi-linearization requirement by establishing a CFDL model and can improve the convergence speed by utilizing historical iterative data.

[0075] In step S3, an encoding-decoding mechanism is established, including:

[0076] Encoder operation: Using logarithmic encoder, according to the formula Perform encoding operations. Where E represents the encoder, is the coded cumulative value at the starting point of the iteration, is the accumulated value of the encoder output, is a dynamic scaling function, is the post-quantization increment.

[0077] The quantization scheme of the logarithmic quantizer Q(·) of the present invention is: Q={±x i :x i =θ i x0,i=0,±1,±2,…}∪{0},0<θ=0.8<1,x0>0 (assuming x0 is a suitable positive number). Dynamic scaling function In each iteration, according to the current calculate Coded cumulative value Implement data encryption and compression to reduce communication load.

[0078] Decoder operation: The decoder is based on the formula Decode the transmitted quantized data back to its original form. Where D represents the encoder, To estimate the output, For the moment No. The estimated output of the iteration.

[0079] It can be deduced that and is a positive number, ensuring decoding accuracy and security), ensuring decoding accuracy and security.

[0080] In step S4, the controller design and pseudo-partial derivative estimate update (the present invention provides an encrypted iterative learning control (EILHC) method) include:

[0081] Controller calculation: According to the formula Calculate the control input. Where, is the target heading angle, ρ is the step size factor, λ is the weight coefficient, is the estimated value of the pseudo partial derivative at the current moment and iteration, Estimated output for the previous iteration.

[0082] In this experiment, the target heading angle is selected as Combined with the current ρ=1、λ=5、 as well as and The value of , determines the control input of each iteration Realize the adjustment of the course of the unmanned surface vessel.

[0083] Pseudo partial derivative estimate update: According to the formula Update the pseudo partial derivative estimates. is the estimated value of the pseudo partial derivative of the previous iteration, η is the learning rate, and μ is the regularization parameter.

[0084] In this experiment, η=1, μ=0.3, and the historical control input increment is used. and estimated output increment Continuous optimization This enables the controller to adjust the control input more accurately.

[0085] In step S5, according to the control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control of the unmanned surface vessel.

[0086] The experiment was carried out by the above embodiment, and the results were as follows Figure 2 As shown. Figure 2 As can be seen from Figures 2(a) and 2(b), the proposed EILHC method outperforms the existing MFAC method in terms of control performance. With appropriate parameter selection, both methods exhibit no significant overshoot, but the EILHC method converges faster, enabling the USV to reach the target heading angle more quickly. Figure 2 (c) and 2(d) demonstrate the effectiveness of the designed encryption scheme. The encoded output is significantly different from the heading angle of the USV. Since the encoder input comes from the redefined virtual output and the encoder output is related to the scaling function, even if the attacker obtains the encoded output, it is difficult to infer the actual state of the USV. For example, if the attacker uses the wrong scaling function Decode from Figure 2 As can be seen from (e) and 2(f), the information obtained is significantly different from the actual USV heading and the information is limited.

[0087] In addition, parameter sensitivity analysis of USV controlled by EILHC method was performed, e.g. Figure 3 shown. Figure 3 (a) shows that increasing ρ from 0 to 1 will shorten the convergence time when other parameters remain unchanged; Figure 3 (b) shows that θ has the least impact on the convergence time, and has greater flexibility in choosing the quantization level; Figure 3 (c) shows that the initial value setting in the quantizer has little effect on the convergence time. It is also found that the parameter selection affects the convergence time and overshoot size. Therefore, in practical applications, it is necessary to select appropriate parameters according to the specific situation to achieve the best control performance.

[0088] Example 2

[0089] like Figure 4 As shown, the present invention provides an iterative learning unmanned surface vehicle heading control system based on data encryption, which is used to implement the iterative learning unmanned surface vehicle heading control method based on data encryption of the above embodiment 1, specifically comprising:

[0090] Sensors for collecting real-time heading angles of unmanned surface vessels and angular velocity

[0091] Redefine the formula Generate Redefinition Output

[0092] Encoder, used to encrypt the redefined output and generate the encoded cumulative value

[0093] Decoder, used to decode the encrypted data and recover the estimated output

[0094] Pseudo partial derivative updater, used to update the pseudo partial derivative estimate based on historical input and output data

[0095] Controller, used to adjust the target heading angle Estimated output and pseudo-partial derivative estimates Generate control input

[0096] Actuator, used to control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control.

[0097] Furthermore, the heading control system in this embodiment consists of a sensor, a redefiner, an encoder, a decoder, a pseudo-partial derivative (PPD) updater, a controller, and an actuator. The sensor is responsible for collecting the heading angle and angular velocity data of the unmanned surface vehicle; the redefiner generates a redefined output based on a formula; the encoder and decoder respectively implement data encryption and restoration; the PPD updater updates the pseudo-partial derivative estimate; the controller calculates the control input based on relevant parameters; and the actuator adjusts the rudder angle of the unmanned surface vehicle based on the control input to achieve heading control.

[0098] The system process is as follows: Sensors collect the heading angle and angular velocity data of the "Dolphin I" unmanned surface vehicle and transmit them to the redefiner to generate a redefined output. The redefined output enters the encoder, undergoes logarithmic quantization and encoding operations, and is transmitted via the communication network. The decoder at the receiving end decodes the encrypted data to obtain the estimated output. The PPD updater updates the pseudo-partial derivative estimate based on historical data. The controller calculates the control input based on the target heading angle, the estimated output, and the pseudo-partial derivative, and sends it to the actuator to adjust the rudder angle, completing a control loop. This process is repeated continuously throughout the experimental time interval to achieve continuous control of the unmanned surface vehicle's heading.

[0099] The present embodiment provides an iterative learning unmanned surface vessel heading control system based on data encryption, which is used to implement the aforementioned iterative learning unmanned surface vessel heading control method based on data encryption. Therefore, the specific implementation method of the iterative learning unmanned surface vessel heading control system based on data encryption can be found in the embodiment section of the iterative learning unmanned surface vessel heading control method based on data encryption above. In order to avoid redundancy, it will not be repeated here.

[0100] Example 3

[0101] An embodiment of the present invention provides a computer storage medium storing a computer software product. The computer software product includes several instructions for enabling a computer device to execute the above-mentioned data encryption-based iterative learning unmanned surface vessel heading control method.

[0102] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0103] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0104] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0105] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will appreciate that other variations or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications derived therefrom remain within the scope of protection of the present invention.

Claims

1. A method for iterative learning of unmanned surface vehicle heading control based on data encryption, characterized in that: include: S1: Redefine the output of the unmanned surface vehicle heading control system as: in, For the moment No. The output of the iteration, For the moment No. The heading angle of the iteration, is the angular velocity at the corresponding moment, K1 is the angular velocity gain; S2: Derivation of output increments through the output redefinition Determine pseudopartial derivatives When satisfied When , a compact form dynamic linearization model is established: Where T is the system time constant, T s is the sampling period, K is the control gain, u max is the maximum input amplitude, Increment for control input; S3: Establish an encoding-decoding mechanism, including: The logarithmic quantizer Q(·) is used to encrypt the redefined output and recursively generate the encoded cumulative value. The formula is: Where E represents the encoder, is the coded cumulative value at the starting point of the iteration, is the accumulated value of the encoder output, is a dynamic scaling function, is the post-quantization increment; The original data is restored based on the coded cumulative value. The formula is: Where D represents the encoder, To estimate the output, For the moment No. The estimated output of the iteration; S4: Based on the compact form dynamic linearized model and the encoded and decoded data, the control input is updated through iterative learning: in, is the target heading angle, ρ is the step size factor, λ is the weight coefficient, is the estimated value of the pseudo partial derivative at the current moment and iteration, Estimated output for the previous iteration; S5: According to the control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control.

2. The iterative learning unmanned surface vehicle heading control method based on data encryption according to claim 1 is characterized in that: The pseudo partial derivative estimate The update formula is: in, is the estimated value of the pseudo partial derivative of the previous iteration, η is the learning rate, and μ is the regularization parameter.

3. The iterative learning unmanned surface vehicle heading control method based on data encryption according to claim 1 is characterized in that: The quantizer Q(·) is defined as: Q={±x i :x i =θ i x0,i=0,±1,±2,…}∪{0},0<θ<1,x0>0; Among them, x i is the quantized value, θ is the quantization base, and x0 is the initial quantization value.

4. The iterative learning unmanned surface vehicle heading control method based on data encryption according to claim 1 is characterized in that: The dynamic scaling function satisfy And dynamically adjust according to the time step or number of iterations, including but not limited to sine scaling function 5. The iterative learning unmanned surface vehicle heading control method based on data encryption according to claim 1 is characterized in that: Parameters satisfy ρ∈(0,1], η∈(0,1], μ>0, λ>λ min , where λ min The minimum weight coefficient is used, and the parameter combination is optimized through sensitivity analysis to ensure that the tracking error is bounded.

6. The iterative learning unmanned surface vehicle heading control method based on data encryption according to claim 1 is characterized in that: The encoder outputs the post-quantization increment With the original heading angle Nonlinear correlation and dynamic scaling function The data compression and encryption are realized by the logarithmic quantizer Q(·), so that the attacker cannot infer the true heading state by intercepting the data.

7. An iterative learning unmanned surface vehicle heading control system based on data encryption, characterized in that: The system is used to implement the unmanned surface vehicle heading control method based on iterative learning of data encryption according to any one of claims 1 to 6, specifically comprising: Sensors for collecting real-time heading angles of unmanned surface vessels and angular velocity Redefine the formula Generate Redefinition Output Encoder, used to encrypt the redefined output and generate the encoded cumulative value Decoder, used to decode the encrypted data and recover the estimated output Pseudo partial derivative updater, used to update the pseudo partial derivative estimate based on historical input and output data Controller, used to adjust the target heading angle Estimated output and pseudo-partial derivative estimates Generate control input Actuator, used to control input Adjust the rudder angle of the unmanned surface vessel to achieve heading control.

8. The iterative learning unmanned surface vehicle heading control system based on data encryption according to claim 7, wherein the logarithmic quantizer Q(·) and the dynamic scaling function in the encoder are Supports dynamic parameter adjustment to adapt to the communication efficiency and security requirements in different marine environments.

9. The method for iterative learning of unmanned surface vehicle heading control based on data encryption according to claim 7, characterized in that: The controller has a built-in parameter adaptive adjustment mechanism to optimize the step size factor ρ and weight coefficient λ in real time to ensure the robustness of the system in complex environments.

10. A computer storage medium, characterized in that The computer storage medium stores a computer software product, which includes several instructions for enabling a computer device to execute the iterative learning unmanned surface vehicle heading control method based on data encryption as described in any one of claims 1 to 6.

Images