Unmanned autonomous underwater vehicle toughness design framework considering multi-source uncertainty

Abstract

The invention provides an unmanned autonomous underwater vehicle (UUV) toughness design framework considering multi-source uncertainty, which comprises four key stages: an initial design stage: optimizing physical parameters and control parameters of a UUV by adopting a multidisciplinary optimization method to obtain initial optimal performance; in the absorption stage, performance changes caused by multi-source uncertainty factors in the task execution process are monitored in real time, the performance degradation condition is analyzed, and the uncertainty type and degree are identified; in the recovery stage, based on the data in the absorption stage, a control strategy is dynamically adjusted through an intelligent optimization algorithm, and the performance of the UUV in the uncertain environment is recovered; in the evolution stage, the uncertainty evolution rule is analyzed through long-term data, physical and control parameters of the UUV are optimized, and the comprehensive performance of the whole application period is improved. According to the method, the environmental adaptability and real-time control capability of the UUV are improved, the design process is optimized, the task endurance time is prolonged, and the method is suitable for the fields of ocean observation, national defense safety, ocean resource development and the like.

Description

Technical Field

[0001] The present invention belongs to the field of ocean exploration technology, and in particular relates to a resilient design framework for unmanned autonomous underwater vehicles taking into account multi-source uncertainties. Background Art

[0002] As humanity explores and develops the ocean, the demand for ocean observation and exploration is growing. Efficient and accurate ocean observation technology is key to in-depth ocean research. Driven by scientific and technological advancements, mobile ocean observation platform technology has made significant progress. Unmanned Underwater Vehicles (UUVs), with their high concealment, low cost, and high reliability, have been widely used in fields such as national defense, security, and ocean development. In recent years, advances in low-power control technology and the rapid development of high-energy-density batteries have significantly increased the endurance of UUVs, enabling them to perform long-term missions in complex ocean environments, further expanding the depth and breadth of ocean observation.

[0003] Current unmanned underwater vehicles (UUVs) face a series of critical issues and shortcomings during long-term missions, severely restricting their practical application effectiveness. First, there is a lack of effective response mechanisms to the combined impact of multiple sources of uncertainty: the dynamic changes in ocean currents lead to reduced navigation accuracy, trajectory deviations, and even a surge in energy consumption for UUVs. Biofouling not only significantly increases navigation resistance and reduces propulsion efficiency, but can also clog sensors or key components, directly impacting data acquisition quality and system reliability. Second, hardware aging and insufficient environmental adaptability: During long-term underwater operations, battery capacity decay, mechanical component wear, and reduced sealing performance exacerbate system performance degradation, while extreme environments further accelerate material fatigue and the risk of electronic component failure, leading to increased mission interruption rates. Furthermore, existing design methods lack dynamic resilience: traditional optimization models are often based on idealized assumptions and struggle to simulate the coupled interactions of multiple factors in complex marine environments, resulting in insufficient robustness in the initial design. Furthermore, real-time monitoring and feedback mechanisms lag behind, making it impossible to quickly identify the type and extent of uncertainty, resulting in delayed control strategy adjustments and limited performance recovery capabilities. More critically, the system lacks the ability to evolve based on long-term data. Most UUVs lack the ability to deeply mine and analyze data from multiple missions, failing to establish a closed-loop link between uncertainty evolution and system parameter updates. This results in design optimization being limited to a single mission and an inability to achieve dynamic performance improvements throughout the entire lifecycle. These issues collectively lead to UUVs facing bottlenecks such as irreversible performance degradation, poor mission adaptability, and high maintenance costs in complex marine environments. Summary of the Invention

[0004] In response to the problems of irreversible performance degradation, reduced system stability and limited life cycle efficiency of unmanned underwater submersibles during long-term missions due to multi-source uncertainties, hardware aging, and insufficient environmental adaptability, the present invention provides a resilient design framework for unmanned autonomous underwater vehicles that takes multi-source uncertainties into consideration.

[0005] The present invention is implemented as follows: a resilient design framework for unmanned autonomous underwater vehicles considering multiple sources of uncertainty, characterized by comprising the following steps:

[0006] (1) Initial design stage: The physical and control parameters of the UUV are optimized through multidisciplinary design optimization methods to obtain the initial optimal performance;

[0007] (2) Absorption phase: During the mission execution of the UUV, the performance changes caused by multi-source uncertainty factors are monitored in real time, performance degradation indicators are recorded and analyzed, and the type and degree of multi-source uncertainty are identified based on the performance degradation situation;

[0008] (3) Recovery phase: Based on the data from the absorption phase, the control strategy of the UUV is dynamically adjusted through an intelligent optimization algorithm to restore its performance in an uncertain environment;

[0009] (4) Evolution stage: By analyzing the uncertainty evolution law in long-term mission data, the physical parameters and control parameters of the unmanned underwater vehicle are redesigned to improve its comprehensive performance throughout the entire application cycle.

[0010] Preferably, in the initial design stage, the multidisciplinary design optimization method takes the coupling relationship between initial physical parameters, control parameters and performance as the optimization target, and solves the optimal initial parameter combination through the intelligent optimization algorithm to obtain the initial optimal performance, that is:

[0011] P0=F0(D0,C0)

[0012] Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P0 represents the initial optimal performance, and F0 represents the coupling relationship among the initial physical parameters, control parameters, and performance.

[0013] Preferably, in the initial design stage, the physical parameters include shape, material, size and energy distribution parameters, and the control parameters include heading angle, attitude angle, navigation trajectory and energy distribution strategy.

[0014] Preferably, in the absorption phase, the performance indicators monitored in real time include speed, energy consumption, trajectory accuracy, and drag coefficient change. The performance degradation indicator is calculated based on the coupling relationship between the initial physical parameters, control parameters, and performance, and the impact of uncertainty on the system is identified through an inverse function, namely:

[0015] P1=F1(D0,C0,U1),R f =P1 / P0×100%.

[0016] U1=F1′(P1,D0,C0)

[0017] Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P1 represents the performance after absorbing uncertainty, F1 represents the complex coupling relationship among physical parameters, control parameters, uncertainty and performance, F1' is the inverse function of F1 used to identify uncertainty, and R f Indicates performance retention rate.

[0018] Preferably, during the recovery phase, the intelligent optimization algorithm is used to solve the optimal control parameters, including real-time adjustment of heading angle, attitude angle, navigation trajectory, and energy allocation strategy. The optimization goal of the recovery phase is to maximize the performance recovery rate, and continuously update the control strategy through a dynamic feedback mechanism. The control strategy and recovery rate are solved as follows:

[0019] maxP1′=F1(D0,C′0,U1)

[0020] R r =(P1′-P1) / P0×100%

[0021] Where C1' represents the optimal control parameter, which can be solved by intelligent optimization algorithm, P1' is the optimal performance in the current system state, R r Indicates the performance recovery rate.

[0022] Preferably, in the evolution stage, the uncertainty evolution law includes the change in drag coefficient and net buoyancy of biofouling; the redesign process establishes a proxy model of physical parameters, control parameters and performance, combines long-term data and a multi-objective optimization algorithm, and solves the global optimal parameter combination, which is expressed as follows:

[0023]

[0024] R e =(P-∑P0) / ∑P0×100%

[0025] Where P represents the overall performance, P i Representing the performance of each profile, based on the set intelligent optimization algorithm of multi-source uncertainty evolution law, with the maximum P as the optimization goal, it can solve the physical parameters D1 and control parameters C that are more suitable for the entire application stage i , R e Indicates the performance improvement rate.

[0026] Preferably, the agent model is established by using an optimized Latin hypercube sampling method and neural network training, with the input being design parameters and uncertainty data, and the output being matching optimal control parameters and performance prediction values.

[0027] Preferably, the multi-source uncertainty includes one or more of ocean current disturbances, biological fouling, system hardware aging and ambient temperature fluctuations.

[0028] Preferably, the resilience design framework achieves dynamic adaptability optimization of the unmanned underwater vehicle in a complex environment by iteratively executing the absorption, recovery and evolution stages.

[0029] This paper proposes a resilient design framework for unmanned autonomous underwater vehicles (UUVs) that considers multiple sources of uncertainty. This framework aims to improve the adaptability, stability, and operational efficiency of UUVs in complex marine environments and under varying mission requirements. This framework not only focuses on optimization strategies during the design phase but also incorporates real-time control mechanisms, enabling the UUV to continuously optimize its performance during operation and ensuring reliable operation under multiple sources of uncertainty.

[0030] First, during the design phase, the present invention employs a process-oriented optimal design approach, comprehensively considering the UUV's physical parameters, control parameters, and multiple sources of uncertainty in the external environment, and analyzing their complex coupling effects on system performance. Traditional design methods often optimize only a single factor, but the present invention overcomes this limitation by introducing a dynamic feedback mechanism that enables the design scheme to be continuously adjusted and optimized based on the interaction of different factors. This approach not only ensures the UUV's strong adaptability during the theoretical design phase but also enables it to more effectively cope with complex marine environments in actual applications, thereby improving overall operational efficiency.

[0031] Secondly, during the mission execution phase, the present invention provides a real-time optimal control mechanism that enables the UUV to dynamically adjust its control strategy based on current environmental conditions. During the mission, the UUV monitors environmental data in real time and assesses the impact of multiple sources of uncertainty (such as ocean current changes, ambient temperature fluctuations, and biological adhesion) on system performance. Based on this data, the system can promptly optimize the control strategy to ensure that the UUV can quickly respond to environmental changes and maintain navigation stability. In addition, through continuous performance evaluation and dynamic adjustment, the control mechanism ensures that the UUV can maximize its recovery and maintain optimal performance when encountering adverse conditions, thereby extending mission execution time and improving mission completion.

[0032] In summary, the advantages of this resilient design framework are reflected in the following aspects: First, it improves the environmental adaptability of UUVs, enabling them to maintain efficient operation in complex and uncertain marine environments; second, it optimizes the design process, improving the rationality and robustness of the design scheme through multi-factor comprehensive analysis and dynamic feedback; third, it enhances the real-time control capability of UUVs, enabling them to autonomously adjust their strategies, improving navigation stability and mission execution capabilities; and fourth, it effectively extends the mission endurance of UUVs, allowing them to maintain a high level of operational performance even during long missions. Overall, the innovation and practicality of this framework will significantly enhance the application value of UUVs in fields such as ocean observation, national defense security, and marine resource development. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a schematic diagram of the internal structure of the underwater glider;

[0034] Figure 2 This is a schematic diagram of the underwater glider's operating mechanism;

[0035] Figure 3 This is a schematic diagram of the overall process of the resilience design framework proposed in this invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0037] To address the irreversible performance degradation, reduced system stability, and limited lifecycle efficiency of unmanned underwater vehicles (UUVs) during long-term missions caused by multiple sources of uncertainty, hardware aging, and insufficient environmental adaptability, this invention provides a resilient UUV design framework that considers multiple sources of uncertainty. To further illustrate the structure of this invention, a detailed description is provided below with reference to the accompanying drawings:

[0038] Underwater glider is a typical underwater unmanned submersible, its internal structure is as follows Figure 1 Its operating principle mainly relies on the synergy of buoyancy adjustment and center of gravity shift: by adjusting its own net buoyancy, it can dive or float; at the same time, by changing the position of the internal battery pack, it can adjust the pitch angle of the glider. The lift generated by the wings in the water not only provides vertical support, but its horizontal component can also push the glider forward, thus achieving efficient underwater navigation. Its specific operating mechanism is as follows Figure 2 The present invention proposes a resilient design framework for unmanned autonomous underwater vehicles that considers multi-source uncertainty, as shown in Figure 3 , including the following steps:

[0039] (1) Initial design stage: The physical and control parameters of the UUV are optimized through multidisciplinary design optimization methods to obtain the initial optimal performance.

[0040] In the initial design phase, the physical parameters include shape, material, size, and energy distribution parameters, and the control parameters include heading angle, attitude angle, navigation trajectory, and energy distribution strategy. The multidisciplinary design optimization method takes the coupling relationship between the initial physical parameters, control parameters, and performance as the optimization target, and uses an intelligent optimization algorithm to solve the optimal initial parameter combination to obtain the initial optimal performance, namely:

[0041] P0=F0(D0,C0)

[0042] Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P0 represents the initial optimal performance, and F0 represents the coupling relationship among the initial physical parameters, control parameters, and performance.

[0043] Specifically, the multidisciplinary design optimization (MDO) method is used to perform the following steps for the underwater glider: Based on previous design experience and the multidisciplinary design optimization method, the basic parameters of the underwater glider are determined, including energy load, shell parameters, space layout, sensor sampling frequency, etc. In addition, the wing parameters are set as the optimizable physical parameters, and the pitch angle and fuel quantity are set as the optimizable control parameters. Therefore, the performance of the underwater glider can be expressed as

[0044] f=F f (W s ,[θ d ,O d ])

[0045] Where f represents specific energy consumption (range / energy consumption), which is a design indicator. s represents the wing parameters, θ d Indicates the pitch angle parameter, O d Indicates the oil quantity parameter.

[0046] Based on the optimal initial performance design goal and intelligent optimization algorithm, the optimal wing parameters and control parameters can be solved;

[0047] Table 1 Initial design results of underwater glider

[0048]

[0049] (2) Absorption phase: During the mission execution of the UUV, the performance changes caused by multi-source uncertainty factors are monitored in real time, the performance degradation indicators are recorded and analyzed, and the type and degree of multi-source uncertainty are identified based on the performance degradation situation.

[0050] During the absorption phase, the performance indicators monitored in real time include speed, energy consumption, trajectory accuracy, and drag coefficient change. The performance degradation indicators are calculated based on the coupling relationship between the initial physical parameters, control parameters, and performance, and the impact of uncertainty on the system is identified through an inverse function, namely:

[0051] P1=F1(D0,C0,U1),R f =P1 / P0×100%.

[0052] U1=F1′(P1,D0,C0)

[0053] Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P1 represents the performance after absorbing uncertainty, F1 represents the complex coupling relationship among physical parameters, control parameters, uncertainty and performance, F1' is the inverse function of F1 used to identify uncertainty, and R f Indicates performance retention rate.

[0054] Through actual sea trials, data was collected to monitor the glider's performance in complex marine environments. The specific steps are as follows: Once the initially designed underwater glider is put into operation, it will be affected by biofouling. Biofouling, as a typical uncertainty, will increase the underwater glider's operating resistance and net buoyancy. A series of sensors, including an electronic compass, depth sensor, CTD, and a cable-attached sensor monitoring device, are deployed on the glider to monitor performance data in real time. The degradation of the underwater glider's performance during the process can be used to identify biofouling.

[0055] [ΔK D0 ,Δ(ΔB)]=F m ′ (P1,D0,C0)

[0056] The contamination status will be used as a reference for real-time control in the next stage. The following table shows the identification results of some typical sections.

[0057] Table 2 Identification results of underwater glider absorption stage

[0058]

[0059]

[0060] (3) Recovery phase: Based on the data from the absorption phase, the control strategy of the UUV is dynamically adjusted through an intelligent optimization algorithm to restore its performance in an uncertain environment.

[0061] The intelligent optimization algorithm is used to solve the optimal control parameters, including real-time adjustment of heading angle, attitude angle, navigation trajectory, and energy allocation strategy. The optimization goal of the recovery phase is to maximize the performance recovery rate, and the control strategy is continuously updated through a dynamic feedback mechanism. The control strategy and recovery rate are solved as follows:

[0062] maxP1′=F1(D0,C′0,U1)

[0063] R r =(P1′-P1) / P0×100%

[0064] Where C1' represents the optimal control parameter, which can be solved by intelligent optimization algorithm, P1' is the optimal performance in the current system state, R r Indicates the performance recovery rate.

[0065] Based on the data from the absorption phase, measures are taken to optimize the glider's control strategy. Based on the performance data from the absorption phase, the glider's heading angle, attitude angle, and energy distribution are adjusted in real time to optimize its performance in uncertain environments. Specific adjustments include adjusting the pitch angle and the amount of fuel used for descents and surfacing to improve recovery performance. The table below shows the real-time status of control parameters and the corresponding performance recovery results.

[0066] Table 3 Control strategy adjustment and recovery effect of underwater glider recovery phase

[0067]

[0068] (4) Evolution stage: By analyzing the uncertainty evolution law in long-term mission data, the physical parameters and control parameters of the unmanned underwater vehicle are redesigned to improve its comprehensive performance throughout the entire application cycle.

[0069] During the evolution phase, the uncertainty evolution law includes the change in drag coefficient and net buoyancy of biofouling. The redesign process establishes a proxy model of physical parameters, control parameters, and performance, combines long-term data, and uses a multi-objective optimization algorithm to solve the global optimal parameter combination, which is expressed as follows:

[0070]

[0071] R e =(P-∑P0) / ∑P0×100%

[0072] Where P represents the overall performance, P i Representing the performance of each profile, based on the set intelligent optimization algorithm of multi-source uncertainty evolution law, with the maximum P as the optimization goal, it can solve the physical parameters D1 and control parameters C that are more suitable for the entire application stage i , R eIndicates the performance improvement rate.

[0073] Specifically, during the evolution phase, data from long-term missions is analyzed to redesign the glider's parameters to improve its overall performance under multi-source uncertainty:

[0074] Evolution law analysis: By analyzing the operational data of multiple missions, we can identify the evolution law of biofouling and build corresponding models to predict its impact on performance. By analyzing the changes in drag coefficient and net buoyancy during several identified missions, we can obtain its evolution law:

[0075] ΔK D0 =1.2 / (1+15×e -0.006×(x-475) )

[0076] Δ(ΔB)=1.2 / (1+10×e -0.006×(x-300) )

[0077] Redesign: Introducing the law of changes in drag coefficient and net buoyancy caused by biofouling into the design, solving the global optimal wing parameters and matching the optimal control parameters for each section. The specific redesign steps are as follows:

[0078] Step 1: Create a proxy model P r , whose inputs are the design parameters and uncertainties, and whose outputs are the matching optimal control parameters. An optimized Latin hypercube sampling method is used to sample and solve for the optimal control parameters. The resulting dataset is then used to build a surrogate model using a neural network. This step aims to simplify the direct solution of complex control parameters and improve optimization efficiency.

[0079] Step 2: Establish a proxy model for performance parameters. Due to its complexity, the performance model obtained in the initial design phase would be computationally intensive if directly applied to the optimization process. Therefore, a proxy model is established and applied to the optimization process. Similarly, the optimized Latin hypercube sampling method is used to sample the design space, solve for the corresponding performance parameters, and establish a proxy model using a neural network. Due to the high computational complexity of the performance model in the initial design phase, this step uses the same sampling and neural network methods to construct a proxy model for performance prediction, replacing the original model for subsequent optimization to reduce computational resource consumption.

[0080] Step 3: Based on the specific energy load, solve the global optimal wing parameters and use the proxy model P r The optimal control parameters for each profile are solved. Under the constraints of a specific energy load, the globally optimal wing parameters (physical parameters) are solved based on the proxy models from steps one and two. The control parameter proxy model is then used to dynamically match the optimal control strategy for each mission profile, achieving comprehensive performance improvements during long-term missions.

[0081]

[0082] Where N is the total number of sections, W s is the wing parameter C si is the control parameter, P r is the proxy model of the optimal control parameters, E T The energy carrying capacity of the underwater glider, and W s U represents the upper and lower limits of the wing parameters, and C s U Indicates the upper and lower limits of the control parameter.

[0083] In this embodiment, the multi-source uncertainty includes one or more of ocean current disturbances, biofouling, system hardware aging, and ambient temperature fluctuations. The proxy model is established using optimized Latin hypercube sampling and neural network training. Its inputs are design parameters and uncertainty data, and its outputs are matched optimal control parameters and performance predictions. The resilient design framework iteratively executes the absorption, recovery, and evolution phases to achieve dynamic adaptive optimization of the unmanned underwater vehicle in complex environments.

[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A resilient design framework for unmanned autonomous underwater vehicles considering multi-source uncertainty, characterized by: The following steps are involved: (1) Initial design stage: The physical and control parameters of the UUV are optimized through multidisciplinary design optimization methods to obtain the initial optimal performance; (2) Absorption phase: During the mission execution of the UUV, the performance changes caused by multi-source uncertainty factors are monitored in real time, performance degradation indicators are recorded and analyzed, and the type and degree of multi-source uncertainty are identified based on the performance degradation situation; (3) Recovery phase: Based on the data from the absorption phase, the control strategy of the UUV is dynamically adjusted through an intelligent optimization algorithm to restore its performance in an uncertain environment; (4) Evolution stage: By analyzing the uncertainty evolution law in long-term mission data, the physical parameters and control parameters of the unmanned underwater vehicle are redesigned to improve its comprehensive performance throughout the entire application cycle.

2. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: In the initial design phase, the multidisciplinary design optimization method takes the coupling relationship between initial physical parameters, control parameters and performance as the optimization target, and solves the optimal initial parameter combination through the intelligent optimization algorithm to obtain the initial optimal performance, namely: P0=F0(D0,C0) Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P0 represents the initial optimal performance, and F0 represents the coupling relationship among the initial physical parameters, control parameters, and performance.

3. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 2 is characterized in that: In the initial design stage, the physical parameters include shape, material, size and energy distribution parameters, and the control parameters include heading angle, attitude angle, navigation trajectory and energy distribution strategy.

4. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: During the absorption phase, the performance indicators monitored in real time include speed, energy consumption, trajectory accuracy, and drag coefficient change. The performance degradation indicators are calculated based on the coupling relationship between the initial physical parameters, control parameters, and performance, and the impact of uncertainty on the system is identified through an inverse function, namely: P1=F1(D0,C0,U1),R f =P1 / P0×100%. U1=F′1(P1,D0,C0) Where D0 represents the initial physical parameters, C0 represents the initial control parameters, P1 represents the performance after absorbing uncertainty, F1 represents the complex coupling relationship among physical parameters, control parameters, uncertainty and performance, F1' is the inverse function of F1 used to identify uncertainty, and R f Indicates performance retention rate.

5. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: During the recovery phase, the intelligent optimization algorithm is used to solve the optimal control parameters, including real-time adjustment of heading angle, attitude angle, navigation trajectory, and energy allocation strategy. The optimization goal of the recovery phase is to maximize the performance recovery rate, and the control strategy is continuously updated through a dynamic feedback mechanism. The control strategy and recovery rate are solved as follows: maxP′1=F1(D0,C′0,U1) R r =(P′1-P1) / P0×100% Where C1' represents the optimal control parameter, which can be solved by intelligent optimization algorithm, P1' is the optimal performance in the current system state, R r Indicates the performance recovery rate.

6. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: During the evolution phase, the uncertainty evolution law includes the change in drag coefficient and net buoyancy of biofouling. The redesign process establishes a proxy model of physical parameters, control parameters, and performance, combines long-term data, and uses a multi-objective optimization algorithm to solve the global optimal parameter combination, which is expressed as follows: R e =(P-∑P0) / ∑P0×100% Where P represents the overall performance, P i Representing the performance of each profile, based on the set intelligent optimization algorithm of multi-source uncertainty evolution law, with the maximum P as the optimization goal, it can solve the physical parameters D1 and control parameters C that are more suitable for the entire application stage i , R e Indicates the performance improvement rate.

7. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 6 is characterized in that: The agent model is established by using an optimized Latin hypercube sampling method and neural network training, with design parameters and uncertainty data as input and the matched optimal control parameters and performance prediction values as output.

8. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: The multi-source uncertainty includes one or more of ocean current disturbances, biological fouling, system hardware aging, and ambient temperature fluctuations.

9. The unmanned autonomous underwater vehicle resilience design framework considering multi-source uncertainty according to claim 1 is characterized in that: The proposed resilience design framework achieves dynamic adaptability optimization of unmanned underwater vehicles in complex environments by iteratively executing absorption, recovery, and evolution stages.

Images