Self-adaptive neural network unmanned ship course control method with input quantization and output constraint

By adopting an adaptive control method based on RBF neural network in the heading control of unmanned ships, combined with input quantization and output constraint theory, the problems of insufficient communication bandwidth and external interference in the marine environment are solved, and the stability of heading control is achieved and security, reducing communication burden and frequency of execution.

CN119045482APending Publication Date: 2024-11-29DALIAN MARITIME UNIVERSITY
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411121733.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

In a complex and changeable marine environment, unmanned ships face heading control challenges caused by insufficient communication bandwidth and position errors. Relying solely on quantitative technology cannot meet the development requirements of unmanned ships, and they need to overcome external interference such as wind, flow, and waves. Avoid collision accidents.

Method used

Adaptive control method based on RBF neural network is adopted, combined with input quantization and output constraint theory, a composite quantizer and Lyapunov function are designed, and the quantization process is described through a linear model, which reduces the execution frequency of the actuator, and ensures the stability of the heading control system and Signal consistency.

Benefits of technology

Effectively reduce the consumption of unmanned ship communication resources, reduce the execution frequency of actuators, ensure the accuracy and safety of heading tracking, and prove the stability of the closed-loop system through the Lyapunov stability theory, avoiding position errors to cause collision risk.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119045482A_ABST
    Figure CN119045482A_ABST
Patent Text Reader

Abstract

The invention provides a self-adaptive neural network unmanned ship course control method with input quantization and output constraint, which comprises the following steps: S1, acquiring sea condition information of a surrounding environment and other surrounding ships, and establishing a course control mathematical model of an unmanned ship; s2, quantizing control input in a control system by adopting a composite quantizer, and describing an input quantization process by utilizing a linear analysis model; s3, designing a barrier Lyapunov function based on an output constraint theory, and obtaining a course controller of the unmanned ship; and S4, based on the Lyapunov stability theory, proving the stability of the designed adaptive neural network unmanned ship course control system with input quantization and output constraint when priori information of a quantization parameter is not needed, and all signals in a closed-loop control system are consistent and finally bounded. According to the technical scheme, the course performance of the unmanned ship can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an adaptive neural network unmanned ship heading control method with input quantization and output constraints. Background Art

[0002] With the development and application of intelligent unmanned ships in marine engineering, the demand for autonomous navigation is becoming increasingly urgent. Unmanned ships have broad application prospects in the fields of marine resource exploration, military mission execution, remote environmental monitoring, etc.

[0003] However, in a complex and changing marine environment, keeping the ship stable and accurately controlling the heading is one of the key challenges to achieve autonomous navigation of unmanned ships. So far, most studies have been conducted to improve the stability and effectiveness of unmanned ship tracking by designing effective and applicable control strategies. Since the components of unmanned ships transmit information through communication channels, and the bandwidth of the communication channels is very limited, unmanned ships face the challenge of insufficient communication bandwidth resources. By using quantization technology before signal transmission, not only can the execution frequency of the actuator be reduced, but also the communication rate within the bandwidth can be reduced. Considering the addition of input quantization technology when studying the heading control problem of unmanned ships has attracted more and more attention.

[0004] Simply relying on quantization technology for heading control cannot fully meet the requirements of the development of unmanned ships. On the one hand, although the use of quantization technology in the control system has certain advantages, the introduction of quantization technology will introduce quantization errors. If these errors are too obvious, it may seriously affect the performance of ship heading tracking. On the other hand, when unmanned ships perform tasks in complex marine environments, they need to overcome challenges such as wind, current, and waves to avoid collision accidents. These environmental factors may limit the position, heading, and speed of unmanned ships. Therefore, position constraints are widely used in heading control systems. Summary of the invention

[0005] According to the above-mentioned technical problems of limited communication bandwidth of unmanned ships at sea and avoiding collisions caused by position errors, the present invention provides an adaptive control method based on RBF neural network to solve the model uncertainty in the control system and the external interference in the heading control subsystem. In addition, a linear model is used to describe the quantization process, which significantly reduces the execution frequency of the actuator, saves communication resources, and reduces the communication burden. Subsequently, an adaptive control law is designed to successfully minimize the tracking error and ensure that the unmanned ship accurately travels along the predetermined route. Finally, the stability of the closed-loop system is proved using the Lyapunov stability theory.

[0006] The technical means adopted by the present invention are as follows:

[0007] An adaptive neural network unmanned ship heading control method with input quantization and output constraints, comprising:

[0008] S1. Obtain the sea condition information of the surrounding environment and other ships around, and establish a mathematical model for the heading control of the unmanned ship;

[0009] S2, using a composite quantizer to quantize the control input in the control system, and using a linear analytical model to describe the input quantization process;

[0010] S3. Based on the output constraint theory, the obstacle Lyapunov function is designed to obtain the heading controller of the unmanned ship;

[0011] S4. Based on Lyapunov stability theory, the stability of the designed adaptive neural network unmanned ship heading control system with input quantization and output constraints is proved when no prior information of quantized parameters is required, and all signals in the closed-loop control system are consistent and ultimately bounded.

[0012] Furthermore, in step S1, the mathematical model of the heading control of the unmanned ship is established as follows:

[0013]

[0014] in, is the ship heading angle, r is the ship's bow angular velocity, n is the control system gain, K, T is the ship maneuverability index; R(r)=c1r+c2r 3 , R(r) is the internal disturbance of the system, c1, c2 are the nonlinear coefficients of the Norrbin motion model, ω represents the unknown disturbance caused by wind, flow, waves and other factors; Q(τ) is the quantized value of the control input τ.

[0015] Further, step S2 specifically includes:

[0016] S21. Use a composite quantizer to quantize the control input in the system. The specific quantization process is described as follows:

[0017]

[0018] Where q1() represents the logarithmic quantizer; f i () is defined as the rounding function less than or equal to (); τ i,s >0 is the switching threshold between logarithmic quantization and uniform quantization; σ i is the quantization density of the uniform quantizer;

[0019] S22. For the quantized control input in the ship model, let Q(τ) = q1(t)τ + q2(t), and take:

[0020]

[0021] Among them, q1(t) is unknown. Since the positive and negative values ​​of q1 do not change during the quantization process, it can be obtained from the above formula that q1(t)>0. When τ(t)<a, ​​Q(τ) is bounded, so q2(t) is bounded.

[0022] S23, convert the heading control mathematical model constructed in S1 into:

[0023]

[0024] Further, step S3 specifically includes:

[0025] S31, define position error, z1 = φ-y d , z2=r-α, α is the stability function to be designed;

[0026] S32. Define the Lyapunov function as follows:

[0027]

[0028] Among them, log() represents the natural logarithm, k b =k c -A0;

[0029] S33, deriving the Lyapunov function defined in step S32, and obtaining:

[0030]

[0031] S34, define the stability function α as Therefore, we have:

[0032]

[0033] S35. Since z2 does not require constraints, define the Lyapunov function as follows:

[0034]

[0035] S36, deriving the Lyapunov function defined in step S35, and obtaining:

[0036]

[0037] S37. Since g(x,t) is unknown, according to the universal approximation theorem, for any small constant ε N , there exists a neural network system W *T h(X), defined as: ΔX = W *Th(X)+ε, let ΔX=g(x,t)+β, where, W * is the ideal weight of the fuzzy logic system, h(X) is the fuzzy basis vector, ε is the approximate error of the fuzzy system and ε≤ε N ;definition Yes W * The estimated value of is:

[0038]

[0039] S38. Design the Lyapunov function as follows:

[0040]

[0041] S39, deriving the Lyapunov function defined in step S38, and obtaining:

[0042]

[0043] S310, define the time-varying gain as follows:

[0044]

[0045] Among them, q1(t) min is the lower bound of q1(t);

[0046] S311. According to step S38 and step S310, the system control law and the adaptive law are designed as follows:

[0047]

[0048] Further, step S4 specifically includes:

[0049] S41, Lyapunov function designed by step S38

[0050] And taking the derivative with respect to V we get:

[0051]

[0052] S42, Order Substituting the control rate τ into the equation, we get:

[0053]

[0054] S43, the adaptive rate and

[0055] Substitute it and get:

[0056]

[0057] S44. Due to Then we have:

[0058]

[0059] make Then we have:

[0060]

[0061] S45, according to Then we get:

[0062]

[0063] S46. Due to:

[0064] Thus we get:

[0065]

[0066] S47. According to the output constraint lemma: for the error system: where η=[w,z1] T , there exist continuously differentiable and positive definite functions V1 and V2, K b >0, the position output is x1, and the position error z1 is defined as x1-y d ,satisfy:

[0067] When Z1→-K a1 Or Z1→K a1 When , V1(z1)→∞;

[0068] γ1(η)≤U(η)≤γ2(η), where γ1 and γ2 are K ∞ Class functions;

[0069] S48. Assume z1(0)<k b , take V(η):=V1(Z1)+U(ω), if it satisfies: Then z1(t)<k b , Then we have: Thus we can get:

[0070]

[0071] S49, Substitute into:

[0072]

[0073] S410, comprehensive consideration, It can be seen that V is bounded, so z1, z2 and They are all bounded. satisfy When t→∞, It converges asymptotically, and the accuracy of convergence depends on η.

[0074] Compared with the prior art, the present invention has the following advantages:

[0075] 1. The present invention provides an adaptive neural network unmanned ship heading tracking control method with input quantization and output constraints. It aims to solve the heading tracking control problem of unmanned ships when the maritime communication bandwidth is limited. It uses quantization technology before signal transmission to save communication resources and reduce the execution frequency of actuators.

[0076] 2. The present invention provides an adaptive neural network unmanned ship heading tracking control method with input quantization and output constraints, which uses a composite quantizer to linearly describe the quantization process, that is, the designed controller does not require prior information of the quantization parameters.

[0077] 3. The adaptive neural network unmanned ship heading tracking control method with input quantization and output constraints provided by the present invention designs the obstacle Lyapunov function according to the output constraint theory, and combines the control rate obtained by the input quantization technology, which can not only reduce the control amplitude of the controller, but also ensure that the controller remains within the safe range when executing the control command. Finally, the effectiveness of the adaptive neural network unmanned ship heading tracking control method with input quantization and output constraints proposed by the present invention is verified through Matlab simulation experiments.

[0078] Based on the above reasons, the present invention can be widely promoted in fields such as artificial intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0080] Figure 1 The figure is a flow chart of the method of the present invention.

[0081] Figure 2 This is a diagram of the tracking results of the ship's heading angle and the ship's bow angular velocity provided in an embodiment of the present invention.

[0082] Figure 3 The tracking error result diagram of the ship's heading angle and the ship's bow angular velocity provided in the embodiment of the invention.

[0083] Figure 4 A comparison diagram of control input curves provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0085] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0086] like Figure 1 The present invention provides an adaptive neural network unmanned ship heading tracking control method with input quantization and output constraints, comprising:

[0087] S1. Obtain the sea condition information of the surrounding environment and other ships around, and establish a mathematical model for the heading control of the unmanned ship;

[0088] S2, using a composite quantizer to quantize the control input in the control system, and using a linear analytical model to describe the input quantization process;

[0089] S3. Based on the output constraint theory, the obstacle Lyapunov function is designed to obtain the heading controller of the unmanned ship;

[0090] S4. Based on Lyapunov stability theory, the stability of the designed adaptive neural network unmanned ship heading tracking control system with input quantization and output constraints is proved when no prior information of quantized parameters is required, and all signals in the closed-loop control system are consistent and ultimately bounded.

[0091] In specific implementation, as a preferred embodiment of the present invention, in step S1, the mathematical model of the heading control of the unmanned ship is established as follows:

[0092]

[0093] in, is the ship heading angle, r is the ship's bow angular velocity, n is the control system gain, K, T is the ship maneuverability index; R(r)=c1r+c2r 3 , R(r) is the internal disturbance of the system, c1, c2 are the nonlinear coefficients of the Norrbin motion model, ω represents the unknown disturbance caused by wind, flow, waves and other factors; Q(τ) is the quantized value of the control input τ.

[0094] In specific implementation, as a preferred embodiment of the present invention, step S2 specifically includes:

[0095] S21. Use a composite quantizer to quantize the control input in the system. The specific quantization process is described as follows:

[0096]

[0097] Where q1() represents the logarithmic quantizer; f i () is defined as the rounding function less than or equal to (); τ i,s >0 is the switching threshold between logarithmic quantization and uniform quantization; σ i is the quantization density of the uniform quantizer;

[0098] S22. For the quantized control input in the ship model, let Q(τ) = q1(t)τ + q2(t), and take:

[0099]

[0100] Among them, q1(t) is unknown. Since the positive and negative values ​​of q1 do not change during the quantization process, it can be obtained from the above formula that q1(t)>0. When τ(t)<a, ​​Q(τ) is bounded, so q2(t) is bounded.

[0101] S23, convert the heading control mathematical model constructed in S1 into:

[0102]

[0103] In specific implementation, as a preferred embodiment of the present invention, step S3 specifically includes:

[0104] S31, define position error, z1 = φ-y d , z2=r-α, α is the stability function to be designed;

[0105] S32. Define the Lyapunov function as follows:

[0106]

[0107] Among them, log() represents the natural logarithm, k b =k c -A0;

[0108] S33, deriving the Lyapunov function defined in step S32, and obtaining:

[0109]

[0110] S34, define the stability function α as Therefore, we have:

[0111]

[0112] S35. Since z2 does not require constraints, define the Lyapunov function as follows:

[0113]

[0114] S36, deriving the Lyapunov function defined in step S35, and obtaining:

[0115]

[0116] S37. Since g(x,t) is unknown, according to the universal approximation theorem, for any small constant ε N , there exists a neural network system W *T h(X), defined as: ΔX = W *T h(X)+ε, let ΔX=g(x,t)+β, where, W * is the ideal weight of the fuzzy logic system, h(X) is the fuzzy basis vector, ε is the approximate error of the fuzzy system and ε≤ε N ;definition Yes W * The estimated value of is:

[0117]

[0118] S38. Design the Lyapunov function as follows:

[0119]

[0120] S39, deriving the Lyapunov function defined in step S38, and obtaining:

[0121]

[0122] S310, define the time-varying gain as follows:

[0123]

[0124] Among them, q1(t) min is the lower bound of q1(t);

[0125] S311. According to step S38 and step S310, the system control law and the adaptive law are designed as follows:

[0126]

[0127] In specific implementation, as a preferred embodiment of the present invention, step S4 specifically includes:

[0128] S41, Lyapunov function designed by step S38 And taking the derivative with respect to V we get:

[0129]

[0130] S42, Order Substituting the control rate τ into the equation, we get:

[0131]

[0132] S43, the adaptive rate and

[0133] Substitute it and get:

[0134]

[0135] S44. Due to Then we have:

[0136]

[0137] make Then we have:

[0138]

[0139] S45, according to Then we get:

[0140]

[0141] S46. Due to:

[0142] Thus we get:

[0143]

[0144] S47. According to the output constraint lemma: for the error system: where η=[w,z1] T , there exist continuously differentiable and positive definite functions V1 and V2, K b >0, the position output is x1, and the position error z1 is defined as x1-y d ,satisfy:

[0145] When Z1→-K a1 Or Z1→K a1 When , V1(z1)→∞;

[0146] γ1(η)≤U(η)≤γ2(η), where γ1 and γ2 are K ∞ Class functions;

[0147] S48. Assume z1(0)<k b , take V(η):=V1(Z1)+U(ω), if it satisfies: Then z1(t)<k b , Then we have: Thus we can get:

[0148]

[0149] S49, Substitute into:

[0150]

[0151] S410, comprehensive consideration, It can be seen that V is bounded, so z1, z2 and They are all bounded. satisfy When t→∞, It converges asymptotically, and the accuracy of convergence depends on η.

[0152] Example

[0153] In order to verify the effectiveness of the solution of the present invention, this embodiment uses MATLAB to perform computer simulation research, and the parameters are set as follows:

[0154] The simulation object is the "Lanxin" unmanned ship of Dalian Maritime University. The parameters of the unmanned ship are as follows: length 7.02m, width 2.6m, speed 35kn, full load draft 0.32m, full load displacement 2.73m 3 , square coefficient 0.6976. Adopt Norrbin motion model, take model parameters K = 0.71, T = 0.32, take nonlinear coefficients a1 = 1, a2 = 0.001.

[0155] Set the ideal heading of the unmanned ship to φ d = sint, the initial state of the actual controlled object is [0,0], and the controller designed in step S3 is used to control the unmanned ship, where the design parameters are as follows: c = 10, l = 30, η = 2, ρ = 0.02, γ1=3,γ2=2, quantization density σ of uniform quantizer i It is 1.42.

[0156] The experimental results are as follows Figure 2-Figure 4 As shown, Figure 2 The tracking results of the unmanned ship's heading angle and bow angular velocity are displayed. It can be seen that the controller designed in the present invention has good control performance and can realize ship heading tracking relatively quickly. Figure 3 is the ship heading angle tracking error and the ship bow angular velocity tracking error. It can be seen that the tracking error can quickly converge to a smaller residual set, which further verifies the effectiveness of the designed controller. Figure 4 This is the control input curve before and after considering input quantization and output constraints. It can be seen from the figure that after comprehensively considering input quantization and output constraints, the execution frequency of the actuator is reduced, and the control system can remain within the safety range when executing commands, ensuring the effectiveness of the control system.

[0157] The simulation results verify the effectiveness of the adaptive neural network controller used in the heading tracking task. The control system can effectively respond to changes and disturbances in the external environment and maintain stable heading control. Considering the input quantization technology, it can not only reduce the communication burden of the unmanned ship, but also reduce the execution frequency of the actuator; the controller designed based on the output constraint theory ensures that the maneuvering signal of the unmanned ship is within a safe range and avoids the adverse consequences that may be caused by excessive or too small output signals; further verifies the effectiveness of the adaptive neural network unmanned ship heading tracking control with input quantization and output constraints proposed in the present invention.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive neural network unmanned ship heading control method with input quantization and output constraints, characterized in that: include: S1. Obtain the sea condition information of the surrounding environment and other ships around, and establish a mathematical model for the heading control of the unmanned ship; S2, using a composite quantizer to quantize the control input in the control system, and using a linear analytical model to describe the input quantization process; S3. Based on the output constraint theory, the obstacle Lyapunov function is designed to obtain the heading controller of the unmanned ship; S4. Based on Lyapunov stability theory, the stability of the designed adaptive neural network unmanned ship heading control system with input quantization and output constraints is proved when no prior information of quantized parameters is required, and all signals in the closed-loop control system are consistent and ultimately bounded.

2. According to claim 1, an adaptive neural network unmanned ship heading control method with input quantization and output constraints is characterized in that: In step S1, the mathematical model of the heading control of the unmanned ship is established as follows: in, is the ship heading angle, r is the ship's bow angular velocity, n is the control system gain, K, T is the ship maneuverability index; R(r)=c1r+c2r 3 , R(r) is the internal disturbance of the system, c1, c2 are the nonlinear coefficients of the Norrbin motion model, ω represents the unknown disturbance caused by wind, flow, waves and other factors; Q(τ) is the quantized value of the control input τ.

3. The method for controlling the heading of an unmanned ship using an adaptive neural network with input quantization and output constraints according to claim 1 is characterized in that: Step S2 specifically includes: S21. Use a composite quantizer to quantize the control input in the system. The specific quantization process is described as follows: Where q1(·) represents the logarithmic quantizer; f i (·) is defined as a rounding function less than or equal to (·); τ i,s >0 is the switching threshold between logarithmic quantization and uniform quantization; σ i is the quantization density of the uniform quantizer; S22. For the quantized control input in the ship model, let Q(τ) = q1(t)τ + q2(t), and take: Among them, q1(t) is unknown. Since the positive and negative values ​​of q1 do not change during the quantization process, it can be obtained from the above formula that q1(t)>0. When |τ(t)|<a, Q(τ) is bounded, so q2(t) is bounded. S23, convert the heading control mathematical model constructed in S1 into:

4. The method for controlling the heading of an unmanned ship using an adaptive neural network with input quantization and output constraints according to claim 1, characterized in that: Step S3 specifically includes: S31, define position error, z1 = φ-y d , z2=r-α, α is the stability function to be designed; S32. Define the Lyapunov function as follows: Where log(·) represents the natural logarithm, k b =k c -A0; S33, deriving the Lyapunov function defined in step S32, and obtaining: S34, define the stability function α as Therefore, we have: S35. Since z2 does not require constraints, define the Lyapunov function as follows: S36, deriving the Lyapunov function defined in step S35, and obtaining: S37. Since g(x,t) is unknown, according to the universal approximation theorem, for any small constant ε N , there exists a neural network system W * T h(X), defined as: ΔX = W *T h(X)+ε, let ΔX=g(x,t)+β, where, W * is the ideal weight of the fuzzy logic system, h(X) is the fuzzy basis vector, ε is the approximate error of the fuzzy system and |ε|≤ε N ;definition Yes W * The estimated value of is: S38. Design the Lyapunov function as follows: S39, deriving the Lyapunov function defined in step S38, and obtaining: S310, define the time-varying gain as follows: Among them, q1(t) min is the lower bound of q1(t); S311. According to step S38 and step S310, the system control law and the adaptive law are designed as follows:

5. The method for controlling the heading of an unmanned ship using an adaptive neural network with input quantization and output constraints according to claim 1, characterized in that: Step S4 specifically includes: S41, Lyapunov function designed by step S38 And taking the derivative with respect to V we get: S42, Order Substituting the control rate τ into the equation, we get: S43, the adaptive rate and Substitute it and get: S44. Due to Then we have: make Then we have: S45, according to Then we get: S46. Due to: Thus we get: S47. According to the output constraint lemma: for the error system: where η=[w,z1] T , there exist continuously differentiable and positive definite functions V1 and V2, K b >0, the position output is x1, and the position error z1 is defined as x1-y d ,satisfy: When Z1→-K a1 Or Z1→K a1 When , V1(z1)→∞; γ1(||η||)≤U(η)≤γ2(||η||), where γ1 and γ2 are K ∞ Class functions; S48. Assume |z1(0)|<k b , take V(η):=V1(Z1)+U(ω), if it satisfies: Then |z1(t)|<k b , Then we have: Thus we can get: S49, Substitute into: S410, comprehensive consideration, It can be seen that V is bounded, so z1, z2 and W are all bounded. Satisfies |z1|<k b , When t→∞, It converges asymptotically, and the accuracy of convergence depends on η.

Citation Information

Cited By

  • Vehicle control method and device

    CN119739085A

  • Intelligent control method and system for operation of rotary air magnetic ship

    CN120010468A

  • Intelligent control method and system for operation of a magnetic ship in rotation

    CN120010468B