A multi-platform observation equipment dynamics system based on marine environment assimilation data

Abstract

A multi-platform observation equipment dynamics system and method based on marine environment assimilation data, comprising a marine environment assimilation data generation module, an equipment deployment area marine dynamic feature extraction module, a multi-scale space environment and force mapping module, a multi-platform observation equipment dynamics time domain solving module, and a multi-body dynamics closed-loop optimization iteration module, wherein the multi-body dynamics closed-loop optimization iteration module feeds back feedback optimization data to the multi-scale space environment and force mapping module, crosses the interdisciplinary barriers of physical oceanography and multi-body dynamics, constructs a standardized data flow conversion, marine dynamic feature extraction, multi-scale force mapping, dynamics solving, artificial intelligence multi-objective optimization closed-loop link, and realizes end-to-end automated intelligent design from marine raw observation data to an optimal equipment structure scheme.

Description

Technical Field

[0001] This invention relates to the field of marine environmental information data processing technology, and in particular to a dynamic system for multi-platform observation equipment based on marine environmental assimilation data. Background Technology

[0002] Marine observation equipment is the core hardware carrier for acquiring three-dimensional environmental parameters such as ocean temperature, salinity, current velocity, biochemistry, and optics. It is the fundamental support for marine scientific surveys, marine environmental monitoring, and underwater safety assurance. Deep-sea mooring buoys, moored buoys, and underwater unmanned observation equipment are deployed long-term in the offshore deep sea, continuously enduring the coupled effects of ocean currents, internal waves, mesoscale eddies, and extreme storm currents. The tension of the mooring cables, the overall drift attitude, and structural deformation of the equipment directly determine the accuracy of the observation data, the service life of the equipment, and the safety of deployment and recovery.

[0003] The current traditional marine observation equipment design system has four major shortcomings: 1. The marine environmental data and equipment design process are disconnected. High-resolution numerical oceanography forecasting and multi-source data assimilation technology can accurately reproduce the three-dimensional multi-scale dynamic processes in the ocean. However, the current equipment design only uses simple empirical flow velocities and has not established a standardized data interface for assimilating environmental fields and dynamic simulations. The load input deviates greatly from the actual marine conditions, and the design is conservative or has potential safety hazards. 2. Lack of universality in equipment design framework. Existing simulation optimization tools are only developed for single types of equipment. The parameter systems, finite element modeling frameworks, and evaluation standards of underwater moorings, buoys, and mobile observation equipment are incompatible with each other. The development of multiple types of equipment requires repeated construction of simulation models, resulting in long development cycles and serious waste of resources. 3. Incomplete coverage of design verification conditions. Traditional designs only use a single average current velocity or a few extreme conditions for verification, without distinguishing between climatological conditions, internal waves, mesoscale eddies, and extreme sea states with different return periods. This results in a large number of design blind spots, which can easily lead to large-scale drift, cable overload breakage, and inaccurate or malfunctioning observation equipment under extreme sea states. 4. The optimization process is highly dependent on manual calculations. After adjustments to the equipment's structural dimensions, buoyancy configuration, and cable segmentation scheme, it is necessary to manually re-import environmental data, model, simulate, and compare the results. This makes it impossible to achieve automatic iterative optimization of parameters, resulting in low optimization efficiency and difficulty in simultaneously balancing the dual indicators of structural safety and observational stability. Summary of the Invention

[0004] To address the problems and technical requirements of existing technologies, the purpose of this invention is to provide a dynamic system for multi-platform observation equipment based on marine environmental assimilation data. This system bridges the interdisciplinary barriers between physical oceanography and multibody dynamics, constructs a closed-loop link for standardized data stream conversion, marine dynamic feature extraction, multi-scale force mapping, dynamic solution, and artificial intelligence multi-objective optimization, and achieves end-to-end automated intelligent design from raw marine observation data to the optimal equipment structure scheme.

[0005] To achieve the above objectives, the present invention employs the following technical solution: A dynamic system for multi-platform observation equipment based on marine environment assimilation data includes a marine environment assimilation data generation module, a marine dynamic feature extraction module for equipment deployment area, a multi-scale spatial environment and force mapping module, a time-domain solution module for multi-platform observation equipment dynamics, and a multi-body dynamics closed-loop optimization iteration module; the multi-body dynamics closed-loop optimization iteration module feeds back the optimized data to the multi-scale spatial environment and force mapping module.

[0006] A dynamics method for multi-platform observation equipment based on marine environmental assimilation data, wherein the dynamics system of the multi-platform observation equipment based on marine environmental assimilation data performs the following steps: Step S1: Acquire multi-source heterogeneous data within the observation area and extract the air-sea boundary forcing conditions. Construct a three-dimensional grid model of the target sea area based on the regional ocean circulation numerical model. Assimilate and fuse multi-element ocean observation data to output a long-sequence standardized ocean environment analysis field dataset with dynamic consistency. Step S2: Receive the sea area coordinate parameters of the observation equipment deployment point, extract the historical three-dimensional ocean current velocity time series of the target sea area from the standardized marine environment analysis field dataset output in Step S1, separate the ocean dynamic signals of different frequency bands through multi-scale bandpass filtering, analyze the current velocity field anomaly signal for vertical spatial projection, and then combine the extreme value statistical distribution to fit the historical maximum current velocity, calculate the theoretical extreme current velocity amplitude of different return periods, and generate a three-dimensional flow field model covering multiple working conditions of extreme dynamic scenarios; Step S3: Using a three-dimensional spatial interpolation method, the macroscopic mesh-scale velocity in the three-dimensional flow field model of the extreme dynamic scenario generated in step S2 is mapped to the microscopic finite element nodes of the parameterized equipment model, and the inflow velocity vector at each node is calculated. Step S4: Construct a unified parameterized dynamic simulation framework compatible with multiple types of observation equipment. Using the core hydrodynamic stability parameters of the observation equipment mapped in Step S3 as load inputs, solve for fluid-structure interaction nonlinear fluid forces and multi-physics coupling loads. Abstract the buoyancy components and equipment compartment of the equipment as rigid body elements and the cables as discrete beam elements. Combine the above multi-physics coupling loads to construct and solve the time-domain dynamic model of the rigid body-discrete beam coupled system, and output the full-time three-dimensional displacement field of the equipment and the cable tension field. Step S5: Construct a dual-objective cost evaluation function that combines the structural safety evaluation function and the observation stability evaluation function. Using a multi-objective genetic algorithm, substitute the full-time three-dimensional displacement field and cable tension field output in Step S4 into the cost function for evaluation. Set safety constraints, iteratively optimize the set of key equipment design parameters, generate a new parameter set, and automatically feed it back to Step S3 for closed-loop recalculation until the Pareto front converges, and output the optimal equipment structural configuration parameter scheme.

[0007] The multi-source heterogeneous data in step S1 includes satellite remote sensing, field measurements, and reanalysis products; multiple elements include temperature, salinity, and flow velocity; and long sequences with dynamic consistency include three-dimensional flow velocity vector fields, temperature and salinity fields, and sea surface wind fields.

[0008] The historical three-dimensional ocean current velocity time series in step S2 is the three-dimensional ocean current velocity time series of the sea area where the deployment point is located; the ocean dynamic signal is the ocean dynamic signal of different frequency bands separated by multi-scale bandpass filtering; the three-dimensional flow field model of extreme dynamic scenarios includes a multi-scenario three-dimensional flow field model covering normal to extreme sea states.

[0009] The observation equipment in step S4 includes deep-sea moorings, observation buoys, and marine unmanned aerial vehicles (UAVs); the core parameters of the hydrodynamic stability of the observation equipment include nodal inflow velocity, sea surface wind field, equipment weight and buoyancy configuration; the multi-physics coupling loads include sea surface wind load, gravity and buoyancy configuration.

[0010] The set of key equipment design parameters in step S5 includes equipment geometry, buoyancy configuration, cable segmentation, and anchorage structure; the safety evaluation function is constructed by weighting the equipment structural strength and attitude stability; the observation stability evaluation function is a dynamic displacement measurement of key detection nodes; and the safety constraint condition is that all cables meet the safety factor threshold constraint.

[0011] Compared with the prior art, the beneficial effects of the present invention are: (1) Break down the barriers of interdisciplinary data flow, realize the standardized coupling of marine assimilation numerical environment field and equipment dynamics simulation, abandon the traditional experience flow velocity input, and make the load boundary conditions fit the real deep-sea multi-scale dynamic environment, greatly reducing the design redundancy and the risk of failure under extreme conditions. (2) Construct a multi-platform universal integrated design framework, with one system compatible with deep-sea moorings, observation buoys, and underwater mobile observation equipment, and unify modeling, simulation, and optimization standards to reduce the workload of repetitive development of multiple types of equipment; (3) Full coverage of design verification conditions, distinguishing between climatological conditions, internal waves, mesoscale eddies, and multi-level return period extreme sea states, eliminating the design blind spots caused by traditional single-condition verification, and fully ensuring the environmental adaptability of equipment throughout its entire life cycle; (4) Fully automatic closed-loop intelligent optimization, based on the NSGA-II multi-objective genetic algorithm, autonomously iterates parameters, simultaneously balancing the two major indicators of structural safety and observational stability, eliminating the need for repeated manual calculations and significantly shortening the equipment research and development design cycle; (5) The full-link quantitative output optimization results can intuitively output the comparison data of equipment drift, cable tension and structural dimensions before and after optimization, and are equipped with three-dimensional attitude visualization simulation, which facilitates engineering implementation, scheme review and equipment production.

[0012] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, the following lists specific implementation methods of the present invention.

[0013] The above and other objects, features and advantages of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings, but this is not intended to limit the invention. Attached Figure Description

[0014] Figure 1 This is a schematic diagram and flowchart of the dynamic system of the multi-platform observation equipment based on marine environmental assimilation data of the present invention; Figure 2 These are flow field input profile feature diagrams for eight typical scenarios of the deep-sea mooring according to the present invention; Figure 3 This is a diagram showing the design state and geometric constraints of the deep-sea mooring without current in this invention. In the diagram, the red dots represent buoyancy modules, the blue dots represent optical spheres, the gray lines represent cables, and the black squares represent anchor blocks. Figure 4 This is a comparison of the three-dimensional morphology of the deep-sea mooring under eight typical scenarios after optimization by this invention. In the figure, the red dots are buoyancy modules, the blue dots are optical spheres, and the black squares are anchor blocks. Detailed Implementation

[0015] To facilitate understanding of the present invention, a more comprehensive description will be provided below. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the present invention.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0017] The specific embodiments provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0018] like Figure 1 As shown, a dynamic system for multi-platform observation equipment based on marine environment assimilation data includes a marine environment assimilation data generation module, a marine dynamic feature extraction module for the equipment deployment area, a multi-scale spatial environment and force mapping module, a time-domain solution module for the dynamics of multi-platform observation equipment, and a multi-body dynamics closed-loop optimization iteration module; the multi-body dynamics closed-loop optimization iteration module feeds back the optimized data to the multi-scale spatial environment and force mapping module.

[0019] A dynamics method for multi-platform observation equipment based on marine environmental assimilation data, wherein the dynamics system of the multi-platform observation equipment based on marine environmental assimilation data performs the following steps: Step S1: The marine environment assimilation data generation module acquires multi-source heterogeneous data within the observation area and extracts air-sea boundary forcing conditions. Based on the high-resolution regional ocean circulation numerical model MITgcm, a three-dimensional grid model of the target sea area is constructed. The multi-element marine observation data is assimilated and fused using the ensemble optimal interpolation (EnOI) method. During the assimilation process, a non-solar heat flux correction strategy is coupled to maintain the dynamic balance between surface heat flux and temperature. Based on the assumption of PV conservation, the vertical structure of temperature and salinity below the boundary layer is dynamically projected and corrected. A long-sequence standardized marine environment analysis field dataset with dynamic consistency is output. ; Step S2: The marine dynamic feature extraction module for the equipment deployment area receives the sea area coordinate parameters of the observation equipment deployment point and extracts the standardized marine environmental analysis field dataset output from Step S1. Extracting historical 3D ocean current velocity time series of the target sea area Ocean dynamic signals of different frequency bands were separated by multi-scale bandpass filtering, and the current velocity field anomaly signal was analyzed and projected vertically into space. Then, the historical maximum current velocity was fitted by the Gumbel extreme value type I statistical distribution to estimate the return period. Theoretical extreme velocity amplitude Generate a three-dimensional flow field model covering extreme dynamic scenarios with multiple operating conditions. ; Step S3: Multi-scale Spatial Environment and Force Mapping Module. To address the spatial scale mismatch between the macroscopic ocean model mesh and the microscopic equipment components, the inverse distance weighted IDW 3D spatial interpolation method is used to achieve force mapping, transforming the 3D flow field model of the extreme dynamic scenario generated in Step S2. The macroscopic model grid point velocity is mapped to the microscopic finite element nodes of the parameterized equipment model, and the high-precision inflow velocity vector of each node is calculated. Eliminate the spatial scale mismatch error between the marine numerical grid and the equipment structure model; Step S4: Multi-platform observation equipment dynamics time-domain solution module. Construct a unified parametric dynamics simulation framework compatible with multiple types of observation equipment. Using the core hydrodynamic stability parameters of the observation equipment mapped in Step S3 as load input, call the Morrison equation to solve the fluid-structure interaction nonlinear fluid force. Incorporating multiphysics coupled loads, the equipment float components and equipment compartment are abstracted as 6-DOF rigid body elements, and the cables are abstracted as large-deformation nonlinear discrete beam elements. Combining the aforementioned multiphysics coupled loads, a time-domain dynamic model of the rigid body-discrete beam coupled system is constructed and solved, outputting the equipment's full-time three-dimensional displacement field. and cable tension field ; Step S5: The multibody dynamics closed-loop optimization iterative module constructs a structural safety evaluation function with the optimization objectives of characterizing structural safety and observational stability. With observation stability evaluation function The dual-objective cost evaluation function is used, and the NSGA-II non-dominated sorting multi-objective genetic algorithm is employed to evaluate the full-time three-dimensional displacement field output in step S4. and cable tension field Substitute the cost function for evaluation, and set the safety constraints to be satisfied by all cables. For the set of key design parameters of the equipment Iterative optimization generates a new parameter set and automatically feeds it back to step S3 for closed-loop recalculation until the Pareto front converges, outputting the optimal equipment structure configuration parameter scheme, simulation verification report, and supporting material list.

[0020] The multi-source heterogeneous data in step S1 includes satellite remote sensing, field measurements, and reanalysis products; multiple elements include temperature, salinity, and current velocity; and long sequences with dynamic consistency include three-dimensional current velocity vector fields, temperature and salinity fields, and sea surface wind fields. The Ensemble Optimal Interpolation (EnOI) method in step S1 performs multi-factor joint data assimilation, and its core state analysis value iterative update process formula is as follows:

[0021] In the formula For the updated three-dimensional marine environment analysis field, For the model's forecast background field, For multi-source observation data vectors, The error covariance matrix of the background field. The measurement noise covariance matrix of the observed data. For the observation operator, This is a localization weighting factor.

[0022] The historical three-dimensional ocean current velocity time series in step S2 is the three-dimensional ocean current velocity time series of the deployment point sea area. The ocean dynamic signals are ocean dynamic signals of different frequency bands separated by multi-scale bandpass filtering; the three-dimensional flow field model of extreme dynamic scenarios includes three-dimensional flow field models covering multiple scenarios from normal to extreme sea states; Among them, the three-dimensional ocean current velocity time series Including the latitudinal direction of the deployment point sea area , via and vertical Historical three-dimensional ocean current velocities in three directions; multi-scale ocean dynamic processes including mesoscale eddies, near-inertial oscillations, and internal wave multi-scale; covering normal to extreme sea states, including extreme events covering normal and recurrence period extreme sea states by scaling the current velocity amplitude; The vertical spatial projection of the analytical velocity field anomaly signal in step S2 is performed using the vertical modal analytical method, and the formula is as follows:

[0023] In the formula For the velocity distribution of a specific dynamic process, This is the time evolution coefficient. This is the analytical solution of the nth vertical standard mode for a specific dynamic process; In step S2, the Gumbel extreme value type I distribution is fitted to the historical maximum velocity to calculate the return period. Theoretical extreme velocity amplitude Its formula is:

[0024] In the formula For position parameters, This is the scale parameter.

[0025] The inverse distance weighted IDW three-dimensional spatial interpolation method in step S3 has the following formula:

[0026] In the formula To equip micro finite element nodes The accurate inflow velocity vector after mapping For nodes With pattern grid points The Euclidean distance between them For the flow velocity of the N nearest pattern grid points, This is the distance attenuation coefficient.

[0027] The observation equipment in step S4 includes deep-sea moorings, observation buoys, and marine unmanned aerial vehicle (UAV) mobile equipment; the core parameters of the hydrodynamic stability of the observation equipment include nodal inflow velocity, sea surface wind field, equipment weight and buoyancy configuration; and the configuration of multiphysics field coupled loads, including sea surface wind load, gravity and buoyancy configuration. The Morrison equation formula in step S4 is:

[0028] In the formula To accurately measure inflow speed and equipment node speed, To ensure precise inflow velocity and equipment node acceleration; The time-domain dynamics model formula for the rigid body-discrete beam coupled system in step S4 is as follows:

[0029] In the formula For the overall quality of the system, For the global damping of the system, The global stiffness matrix of the system. For wind load, For buoyancy load, The load is due to gravity.

[0030] The set of key equipment design parameters in step S5 includes equipment geometry, buoyancy configuration, cable segmentation, anchoring structure, etc. The safety evaluation function in step S5 is constructed by weighting the equipment structural strength and attitude stability. The security evaluation function in step S5 The formula is:

[0031] In the formula For the set of key design parameters of the equipment, For the first The actual maximum tension of the cable segment The inherent breaking tension; The observation stability evaluation function in step S5 is the dynamic displacement metric of the key detection node. The observation stability evaluation function in step S5 The formula is:

[0032] In the formula , , The three-dimensional dynamic displacement of the target detection node; The safety constraint in step S5 is that all cables meet the safety factor threshold constraint. The safety constraint in step S5 is that all cables meet the safety factor threshold. The formula is:

[0033] In the formula, This is a preset safety factor.

[0034] The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method.

[0035] Example 1 Dynamic optimization design of deep-sea observation mooring system: This embodiment focuses on the dynamic optimization design of a deep-sea observation station in the central South China Sea.

[0036] 1. Steps S1-S2: Generation of marine environmental assimilation data and extraction of marine dynamic characteristics of equipment deployment area: In step S1, the marine environment assimilation data generation module first collects and integrates multi-source heterogeneous observation data of the sea area. Specifically, it includes: (1) satellite remote sensing data: AVISO sea surface altimeter data (1993-2020, weekly average, resolution 1 / 4°), SST sea surface temperature data (resolution 1 / 4°); (2) field measured data: historical CTD (temperature, salinity, depth) profile observations of the sea area and adjacent areas, ADCP (acoustic Doppler current profiler) fixed-point current velocity observations, Argo buoy temperature, salinity profile data; (3) reanalysis products: ERA5 sea surface atmospheric forcing field (10-meter wind speed, sea surface pressure, shortwave radiation, longwave radiation, precipitation, specific humidity, etc., hourly, resolution 0.25°). Based on the above multi-source heterogeneous data, the system uses the high-resolution regional ocean circulation numerical model MITgcm to construct a three-dimensional grid model of the target sea area. The horizontal direction uses a spherical coordinate grid with a resolution of 1 / 12° (approximately 9 km), covering the range of 12°N–18°N and 112°E–118°E. The vertical direction uses 50 z-coordinate layers, with the top 10 layers having a resolution of 5 m, gradually increasing to 200 m towards the bottom, with the bottom layer located at a depth of 3800 m. Open boundary conditions are driven daily by the global HYCOM reanalysis product, and sea surface forcing is achieved using the ERA5 hourly atmospheric field.

[0037] The Ensemble Optimal Interpolation (EnOI) method was used for the joint assimilation of multi-factor temperature, salinity, and flow data. The ensemble size was set to 100 members, and the initial ensemble sample was generated by randomly perturbing the historical data ensemble. The localization scheme employed a Gaspari-Cohn fifth-order piecewise rational function, with a horizontal localization radius of 150 km and a vertical localization radius of 300 m. The observation operator mapped the model state variables (temperature, salinity, and velocity) to the observation space, and the background error covariance matrix was statistically estimated from the ensemble sample and locally corrected. The assimilation period was 7 days. During the assimilation process, a non-solar heat flux correction strategy was coupled to maintain the dynamic balance between surface heat flux and temperature. Based on the assumption of PV conservation, the vertical temperature, salinity, and flow structure below the boundary layer was dynamically projected and corrected to ensure the hydrostatic stability and geostrophic balance of the assimilated 3D temperature, salinity, and flow field. The final output was a daily 3D temperature, salinity, and flow normalized assimilation data field dataset spanning 28 years (1993-2020).

[0038] In step S2, the marine dynamic feature extraction module of the equipment deployment area receives the precise coordinates of the mooring deployment point and extracts the 28-year daily three-dimensional ocean current velocity time series matrix V(u,v,w,t) of the point from the above-mentioned assimilation data field dataset, where u is the latitudinal component (positive to the east), v is the meridional component (positive to the north), and w is the vertical component (positive upward), with a time resolution of 1 day and a vertical coverage of 50 depths.

[0039] Multi-scale dynamic signal separation was performed on the original velocity time series. Butterworth digital bandpass filter was used to extract three main types of dynamic signals: (1) mesoscale eddy process: the filter frequency band is 30-120 days, corresponding to the typical time scale of Rossby waves and mesoscale eddies. After extraction, the velocity standard deviation was calculated to characterize the intensity of mesoscale disturbance; (2) near-inertial oscillation: the filter frequency band is 1.5-5 days, corresponding to the typical period of near-inertial internal wave in the f plane (the inertial period at this latitude is about 3.8 days). After extraction, its vertical shear characteristics were analyzed; (3) internal wave process: the filter frequency band is 0.5-1.5 days, corresponding to the typical time scale of high-frequency internal wave. In addition, the low-frequency component (>120 days) of the original sequence was retained as the climatological background flow field. Vertical mode decomposition was performed on the velocity field of each scale after filtering and separation. The three-dimensional velocity anomaly signal was projected onto the vertical standard mode basis function to obtain the time evolution coefficient of each mode. This spatial projection effectively extracts the vertical structural features of processes at various scales, providing physical constraints for subsequent extreme flow field modeling.

[0040] Based on the climatological background flow field and disturbance components at various scales, the annual maximum three-dimensional flow velocity is fitted using the Gumbel extreme value type I distribution. The specific process is as follows: For the daily three-dimensional flow velocity time series over 28 years, maximum flow velocity amplitude samples are extracted annually, resulting in 28 annual extreme value samples; the location parameter μ and scale parameter σ of the Gumbel distribution are fitted using the maximum likelihood estimation method (MLE); the goodness of fit is verified through the Kolmogorov-Smirnov test (significance level α=0.05, KS statistic D=0.118 in this embodiment, p=0.732, passing the test). Based on the fitted Gumbel distribution, the theoretical maximum flow velocity amplitude V_return for different return periods Tr is calculated. Finally, three-dimensional flow field models V_extreme(Z) for eight typical extreme dynamic scenarios are generated, covering a complete range of operating conditions from normal to 100-year return period extreme sea states. The relationship between the vertical maximum flow velocity and depth for each scenario is shown below. Figure 2 As shown.

[0041] 2. Steps S3-S4: Multi-scale spatial environment and force mapping, and time-domain solution of dynamics for multi-platform observation equipment: In step S3, the multi-scale spatial environment and force mapping module employs the inverse distance weighted IDW three-dimensional spatial interpolation algorithm to map the macroscopic ocean model grid-scale velocity field generated in step S2 to the microscopic finite element nodes of the parameterized equipment model in step S4. Specifically, for each finite element node i in the mooring structure model, the system searches for the N=8 nearest grid points in the MITgcm model grid, calculates the three-dimensional Euclidean distance d_ij from each grid point to the node, and calculates the weight w_ij=d_ij^(-p) with a distance decay exponent p=2.0. The inflow velocity vector V_node_i at node i is obtained by weighted averaging of the velocities V_j of each neighboring grid point. This interpolation scheme effectively eliminates the spatial scale mismatch error between the ocean numerical grid (resolution approximately 9 km) and the equipment structure model (node ​​spacing approximately 1-5 m), ensuring the accuracy of the load boundaries.

[0042] In step S4, a unified parametric dynamic framework for the deep-sea mooring is constructed. The initial design parameters are as follows: the mooring system is 800 m in total length, the main cable is a Φ16 mm stainless steel wire rope (breaking strength approximately 45 kN), and it has three observation equipment compartments (located at depths of 200 m, 400 m, and 600 m respectively), with a 300 kg gravity anchor at the bottom. The buoyancy system uses two 17-inch glass buoys (each with a net buoyancy of approximately 35 kg), distributed at the top and middle positions. The top equipment compartment carries an optical backscatter sensor and a temperature, temperature, and depth (CTD) meter, requiring the sensor's horizontal pointing stability to be within ±5°.

[0043] The Morrison equations are used to calculate the nonlinear fluid force loads at each node. The fluid density ρ_w is taken as 1025 kg / m³, the drag coefficient C_D is obtained from a table based on the cable diameter and Reynolds number (C_D=1.2 for laminar flow, C_D=0.7 for turbulent flow, and the average value C_D=0.95 in this embodiment), the inertia coefficient C_M is taken as 2.0 (theoretical value for a cylinder), and the windward projected area A_i of the node is automatically calculated from the equipment geometry parameters. The inflow velocity V_node_i of each node and the equipment node velocity and acceleration are substituted into the Morrison equations to obtain the nonlinear fluid force F_fluid_i of each node.

[0044] In terms of structural discretization, the buoy and equipment compartment are abstracted as 6-DOF rigid body elements, each containing 3 translational and 3 rotational degrees of freedom. The main cable is discretized as 100 nonlinear large deformation beam elements. Each beam element adopts Timoshenko beam theory to consider shear deformation effects, and the material constitutive model adopts a bilinear elastoplastic model (elastic modulus E = 200 GPa, yield strength σ_y = 520 MPa). The global system contains a total of 303 nodes and 606 degrees of freedom. Multi-field coupled loads include: sea surface wind load (calculated according to API specifications, wind speed at 10 meters height provided by ERA5 data, wind pressure coefficient taken as 1.2), gravity load (considering the self-weight of each component and the effect of added mass), and buoyancy configuration load (calculated from the displacement of the glass buoy, considering the change of hydrostatic pressure with depth). The Newmark-β implicit time integration method was used to solve the time-domain dynamic differential equations of the coupled system, with a time step Δt = 0.05 s (satisfying the CFL stability condition), a total simulation duration of 3600 s (covering a complete cycle of a typical mesoscale process), and an output time resolution of 1 s. In the initial design scheme, during a mesoscale process simulation with a flow velocity of 0.1 m / s, the top equipment compartment experienced a horizontal drift exceeding 150 meters, a vertical settlement of 12 m, and an inclination angle exceeding 18°, severely exceeding the ±5° pointing stability requirement of the optical sensors. The maximum tension in the main cable reached 38.6 kN, approaching 85% of the breaking strength (insufficient safety margin). The middle section of the cable exhibited a localized large-curvature bend of approximately 2.3 m, posing a risk of entanglement. The system automatically triggered a closed-loop optimization iteration of multibody dynamics.

[0045] 3. Step S5, multibody dynamics closed-loop optimization iteration: The multibody dynamics closed-loop optimization iterative module constructs a dual-cost evaluation function using the structural safety evaluation function f_T(Θ) and the observation stability evaluation function f_X(Θ), and employs the NSGA-II non-dominated sorting genetic algorithm for multi-objective optimization. The optimization parameters are set as follows: population size NP=80, maximum number of iterations G_max=150, crossover probability P_c=0.9 (using simulated binary crossover SBX, distribution exponent η_c=20), and mutation probability P_m=0.15 (using polynomial mutation, distribution exponent η_m=20). The Pareto front convergence criterion is set as follows: convergence is determined when the rate of change of the hypervolume index of the Pareto front is less than 0.1% for 10 consecutive generations.

[0046] The key design parameter set Θ of the equipment includes 6 optimization variables: (1) total length of the mooring L (range 500-900 m, continuous variable); (2) diameter of the main cable D_cable (range Φ12-Φ20 mm, discrete variable, step size 2 mm); (3) number of glass buoys N_buoy (range 2-8, integer variable); (4) buoy specifications S_buoy (selectable 17-inch / 24-inch / 28-inch specifications, combination variable); (5) spacing between observation equipment compartments Δz (range 20-80 m, continuous variable); (6) weight of bottom anchor block W_anchor (range 200-800 kg, continuous variable). The safety constraint is set as the maximum tension ratio of all cables f_T(Θ)≤1 / 3.0 (safety factor S≥3.0), and the observation stability constraint is set as the maximum dynamic displacement of the top equipment compartment does not exceed 80 m.

[0047] During the optimization process, each individual in each generation of the population corresponds to a complete design scheme. The system first feeds back the design scheme parameters Θ to the multi-scale spatial environment and force mapping module, automatically updates the finite element model (adjusting cable length, node coordinates, buoyancy distribution, etc.), and then calls the dynamic time-domain solver in step S4 to calculate the dynamic response of the scheme under 8 typical scenarios, extracts the time-series displacement field X(t) and cable tension field T(t), and substitutes them into the dual-cost function for evaluation. Among them, the safety evaluation function f_T(Θ) takes the maximum value of the ratio of tension to breaking tension of all cable segments under all scenarios, and the observation stability evaluation function f_X(Θ) takes the root mean square value of the three-dimensional dynamic displacement of the top equipment cabin under all scenarios. After 87 generations of iteration, the system determines that the Pareto front has converged (the rate of change of Hypervolume is less than 0.05% for 10 consecutive generations). Finally, the compromise optimal solution is selected from the converged Pareto front solution set as the recommended scheme. This scheme satisfies the safety constraints in all 8 scenarios and has the best observation stability index under normal working conditions.

[0048] The optimized final design parameters are as follows: the total length of the mooring system has been reduced from the initial 800 m to 650 m, a reduction of 18.75%, effectively reducing the current-receiving area; the main cable has been upgraded to a Φ18 mm high-strength stainless steel wire rope (breaking strength 52 kN); the buoyancy system has been reconfigured into four 28-inch high-performance deep-sea glass buoys (net buoyancy of 93.275 kg per buoy, total net buoyancy of 373.1 kg), respectively arranged at the top (2 in parallel), at a depth of 300 m (1), and at a depth of 500 m (1), forming a gradient buoyancy configuration; the spacing between the observation equipment compartments adopts a gradient layout strategy (50 m in the first section, equidistant 30 m in the subsequent sections), optimizing the hydrodynamic shape; the bottom anchoring system has been upgraded to a 600 kg gravity anchor block combined with a Type 8242 parallel dual acoustic release device, achieving a near 100% success rate in safe recovery. A schematic diagram of the mooring structure is shown below. Figure 3 As shown.

[0049] Comparison of design scheme output and closed-loop verification. The system outputs the final optimized scheme. Through parameter optimization, the structural parameters and current-resistant dynamic response of the underwater glider system achieve a perfect closed loop. Table 1 shows a comparison of the technical parameters of the underwater glider structure before and after optimization, as well as key response indicators under typical environments. Table 1 Configuration of Dynamics Optimization Results for Deep-Sea Observation Mooring System

[0050] 4. Closed-loop verification effect: like Figure 4As shown, the optimized 650 m design parameters were substituted back into the system, and a full-coverage closed-loop verification was conducted under eight typical scenarios. In the climatological average scenario (maximum current velocity 0.0068 m / s), the horizontal drift of the mooring tip was 3.2 m, the maximum cable tension ratio was 0.08, and the system was in an excellent stable state. In the extreme mesoscale eddy scenario (maximum current velocity 0.18 m / s), the tip drift converged to 9.8 m, and the maximum tension ratio was 0.42, far below the safety factor threshold. In the 50-year return period extreme sea state (maximum current velocity 0.2136 m / s), the tip displacement was 38.5 m, the maximum cable tension ratio was 0.68, the structure maintained a smooth catenary morphology, and there was no local buckling. Under the most stringent 100-year return period extreme sea state (maximum current velocity 0.2673 m / s), the horizontal displacement of the mooring tip strictly converged to 67.2 m, far below the constraint limit of 80 m; the maximum cable tension ratio was 0.81 (corresponding to a tension of 42.1 kN and a safety factor of 1.23), meeting the equivalent verification requirement of the safety constraint condition f_T(Θ)≤1 / 3.0; the cable maintained a smooth catenary shape throughout the entire process, without any folding, buckling, or entanglement; the array geometry did not become unstable, and all four buoy groups maintained positive buoyancy, effectively suppressing the overall sinking trend. Closed-loop verification results show that the optimized mooring system meets the dual requirements of structural safety and observation stability in all eight typical scenarios.

[0051] A comparison of core performance indicators before and after optimization shows that: horizontal drift at normal current speeds (0.1 m / s) has been significantly reduced from over 150 m to 12.6 m (an improvement of 91.6%); the tip displacement under extreme sea conditions (once in 100 years) has been reduced from a risk of instability to a controllable 67.2 m; the maximum cable tension ratio has been reduced from nearly dangerous 0.85 to a safe level below 0.81; and the total system length has been reduced by 18.75% while buoyancy performance has been improved by approximately 247%. The optimization scheme significantly improves observation accuracy and system reliability without sacrificing safety margins.

[0052] Example 2 Dynamic optimization design of shallow-sea observation buoys: This embodiment focuses on the dynamic optimization design of a shallow-sea mooring buoy system at an observation station in the northern continental shelf area of ​​the South China Sea.

[0053] 1. Steps S1-S2: Generation of marine environmental assimilation data and extraction of marine dynamic characteristics of equipment deployment area: In steps S1-S2, the marine environment assimilation data generation module integrates multi-source observation data of the sea area and extracts extreme flow field information of buoy deployment points from the assimilation data field. Since the deployment sea area is the same as in Example 1, this part will not be described in detail. The maximum flow velocity and vertical distribution characteristics of each scenario are also described.

[0054] 2. Steps S3-S4: Multi-scale spatial environment and force mapping, and time-domain solution of dynamics for multi-platform observation equipment: In step S3, the IDW three-dimensional spatial interpolation algorithm is used to map the macroscopic model mesh velocity field to the nodes of the buoy finite element model. The buoy system adopts a disc-shaped floating structure with a diameter of 2.6 m, a draft of 0.8 m, and a total displacement of approximately 4.2 tons. The mooring system is a catenary single-point mooring system, initially designed with a single Φ20 mm nylon cable (120 m in length, linear density of 1.8 kg / m, elastic modulus E = 2.5 GPa), connected to a 400 kg holding anchor at the bottom. For the 126 finite element nodes in the buoy structural model, the nearest N=8 MITgcm mesh points are searched for IDW interpolation, with a distance attenuation index p = 2.5 (considering the large spatial gradient of the current field in the continental shelf area where the buoy is located, the attenuation index is appropriately increased to improve local resolution).

[0055] In step S4, a unified parametric dynamic framework for the shallow-sea buoy is constructed. The buoy adopts a 6-DOF rigid body model, considering nonlinear hydrodynamic coefficients: vertical additional mass coefficient C_ah=0.8, roll additional moment of inertia coefficient C_aphi=0.35, pitch additional moment of inertia coefficient C_atheta=0.35, and vertical damping coefficient C_dh=2.5. The mooring cable is discretized into 60 Timoshenko beam elements, considering the geometric nonlinearity (large deformation) and material nonlinearity (creep effect of nylon cable, using a three-parameter Maxwell viscoelastic model, relaxation time τ=3600 s) of the cable under tension.

[0056] The Morrison equations were used to calculate the hydrodynamic loads on the floating body and mooring system. The fluid density ρ_w was taken as 1024 kg / m³ (considering the stratification effect of the continental shelf area; surface density 1022 kg / m³, bottom density 1027 kg / m³, linearly interpolated by depth). The floating body drag coefficient C_D was taken as 1.0 (an empirical value for disc-shaped structures), the inertia coefficient C_M as 1.8, and the cable drag coefficient C_D as 1.1. Sea surface wind loads were calculated according to the API RP 2D standard, and the drag coefficient varied with wind speed according to the Charnock relationship.

[0057] The time integration employed the Newmark-β method (β=0.25, γ=0.5, ensuring unconditional stability), with a time step Δt=0.02 s (the buoy system's natural period is approximately 8 s, with 400 samples per period), resulting in a total simulation duration of 7200 s. The initial design's dynamic response under a 0.15 m / s mesoscale eddy condition was as follows: the buoy drifted horizontally by over 85 m, the mooring cable's maximum tilt angle reached 42°, the tip tension fluctuation amplitude reached ±35%, and the buoy's roll angle exceeded 12°, significantly exceeding the stability specifications of the onboard optical sensor (requiring a horizontal pointing error <±3° and a tilt angle <±5°). The system determined that the initial design did not meet the observation requirements and triggered parameter optimization.

[0058] 3. Step S5, multibody dynamics closed-loop optimization iteration: The multibody dynamics closed-loop optimization iterative module constructs a dual-cost evaluation function using the structural safety evaluation function f_T(Θ) and the observation stability evaluation function f_X(Θ), and employs the NSGA-II algorithm for multi-objective optimization. Optimization parameter settings: population size NP=100, maximum number of iterations G_max=200, crossover probability P_c=0.85 (SBX distribution exponent η_c=15), mutation probability P_m=0.12 (polynomial mutation distribution exponent η_m=15). Pareto front convergence criterion: the rate of change of Hypervolume is less than 0.05% for 15 consecutive generations.

[0059] The key design parameter set Θ of the equipment includes 7 optimization variables: (1) Buoy diameter D (range 2.4-3.5 m, continuous variable); (2) Buoy hull line parameters (selectable disc / dish / hemispherical type, combined variable); (3) Mooring cable type (selectable nylon cable / polyester cable / steel wire rope, combined variable); (4) Mooring cable diameter D_cable (range Φ16-Φ28mm, discrete variable); (5) Mooring method (selectable single-section / two-section / three-section counterweight, combined variable); (6) Total mooring length L_moor (range 80-150 m, continuous variable); (7) Bottom anchor weight W_anchor (range 300-600kg, continuous variable). Safety constraints: maximum cable tension ratio f_T(Θ)≤1 / 2.5 (safety factor S≥2.5), maximum buoy tilt angle<15°.

[0060] During the optimization process, a hybrid encoding strategy (real-number encoding of continuous variables + integer encoding of discrete variables) is adopted for combined variables. Each individual evaluation requires the complete execution of the dynamic simulation process in steps S3-S4. Generations 1-50 are the rapid exploration phase, focusing on searching the feasible region boundary; generations 51-120 are the fine development phase, exploring Pareto optimal solutions in depth within the discovered feasible regions; generations 121-200 are the convergence verification phase, verifying the stability of the frontier solution set. After 143 iterations, the system determines that the Pareto front has converged.

[0061] The final optimization scheme selected a compromise solution: the buoy diameter was increased from 2.6 m to 3.0 m to increase the current-receiving area and improve stability, while reducing roll response; the buoy profile was optimized from a disc shape to a saucer shape (bottom curvature radius 1.8 m) to effectively reduce heave amplitude; a streamlined fairing (0.3 m high, 30° cone angle) was added to the bottom to suppress vortex-induced vibration; the mooring system was upgraded to a Φ24 mm polyester cable (elastic modulus E=6.5 GPa, breaking strength 68 kN), adopting a three-section counterweight design (no counterweight in the upper 30 m, uniform counterweight of 5 kg / m in the middle 40 m, and concentrated counterweight of 150 kg in the lower 25 m) to make the tension more evenly distributed along the cable length; the total mooring length was optimized from 120 m to 95 m, shortening it by 21% to reduce the drift radius; and the bottom anchor weight was adjusted to 450 kg.

[0062] Table 2 shows a comparison of the structural technical parameters of the buoy system before and after optimization, as well as typical environmental response indicators.

[0063] Table 2 Configuration table of dynamic optimization results for shallow sea observation buoy system

[0064] 4. Closed-loop verification effect: Under climatological average conditions (maximum flow velocity 0.0085 m / s), the buoy drifted horizontally by 1.8 m, with a mooring line tilt angle of only 4.2°, indicating excellent system stability. Under conventional mesoscale eddy conditions (maximum flow velocity 0.15 m / s), the horizontal drift converged to 7.3 m, the mooring line tilt angle was 11.5°, and the buoy tilt angle was 3.8°, all meeting sensor stability requirements. Under extreme internal wave conditions (maximum flow velocity 0.22 m / s), the horizontal drift was 18.6 m, and the maximum mooring line tilt angle was 15.2°. Due to the buffering effect of the three-section counterweight, the peak tension only reached 48% of the breaking strength.

[0065] Under the most stringent 100-year return period extreme sea state (maximum current velocity 0.3126 m / s), the buoy's horizontal drift strictly converged to 42.3 m, the maximum mooring cable tilt angle was 24.6° (below the 30° safety threshold), the maximum tension ratio was 0.71 (safety factor 1.41), and the maximum float tilt angle was 8.7° (below the 15° constraint). The structure remained stable without resonance or galloping. The verification results under the 50-year return period extreme sea state (maximum current velocity 0.2784 m / s) were even more ideal: horizontal drift of 31.6 m, cable tilt angle of 18°, tension ratio of 0.58, and float tilt angle of 6.2°, meeting the dual requirements of ±3° horizontal pointing and ±5° tilt angle of the optical sensor.

[0066] Comparison of core performance before and after optimization: Under normal current velocity (0.15 m / s), horizontal drift was significantly reduced from over 85 m to 7.3 m (an improvement of 91.4%), and mooring cable inclination angle was reduced from 42° to 11.5° (a reduction of 72.6%); under extreme sea conditions lasting 50 years, the tip displacement was reduced from the risk of instability to a controllable level of 31.6 m; the tension fluctuation amplitude of the mooring system was reduced from ±35% to ±12%, significantly reducing fatigue damage accumulation. The optimization scheme, by increasing the buoy's center of gravity height, optimizing the mooring stiffness distribution, and introducing a counterweight buffer mechanism, achieved high-precision observation capabilities for shallow-sea buoys in complex continental shelf dynamic environments.

[0067] Example 3 Dynamic optimization design of a multi-platform collaborative observation system: This embodiment focuses on the dynamic optimization design of a multi-platform collaborative observation system in the equatorial region of the Indian Ocean.

[0068] This sea area is located at the intersection of the equatorial countercurrent zone and the monsoon circulation. It is modulated by multiple scale dynamic processes, including the equatorial current system (South Equatorial Current SEC, Equatorial Countercurrent EUC, North Equatorial Countercurrent NECC), seasonal circulation reversal driven by the monsoon, Madden-Julian oscillation (MJO, period 30-60 days), and high-frequency equatorial waves (Kelvin waves, Rossby waves). The hydrodynamic environment exhibits significant equatorial characteristics—strong vertical shear, low-frequency large-amplitude waves, and high-frequency internal waves superimposed. This places extremely high demands on the structural design and collaborative deployment of multi-platform collaborative network observation equipment.

[0069] 1. Steps S1-S2: Generation of marine environmental assimilation data and extraction of marine dynamic characteristics of equipment deployment area: In step S1, the marine environment assimilation data generation module integrates multi-source observation data for the sea area. The data sources include: (1) satellite remote sensing data: AVISO sea surface altimeter data (1993-2020), SST / SSS (sea surface temperature / salinity) data, ASCAT sea surface wind stress data (resolution 0.25°, daily); (2) field measured data: long-term observation data of RAMA buoy array in the sea area, historical CTD profile data of WOCE / CLIVAR global ocean circulation experiment, and current velocity observation data of the autonomous mooring array in the equatorial Indian Ocean; (3) reanalysis products: ERA5 hourly sea surface atmospheric forcing field, SODA ocean reanalysis products.

[0070] A high-resolution 3D mesh model of the target sea area was constructed using MITgcm. The horizontal resolution was 1 / 12° (approximately 9 km × 9 km in the equatorial region), covering the area from 5°S to 5°N and from 75°E to 85°E. Vertically, 55 z-coordinate layers were used: the surface 20 layers with a resolution of 5 m (the equatorial thermocline is typically located at a depth of 80-150 m, requiring more refined vertical resolution for accurate characterization); the middle 25 layers with a resolution of 15 m; the bottom 10 layers with a resolution of 100 m; and the bottom layer located at 4300 m. Ensemble optimal interpolation (EnOI) assimilation parameters were: 120 ensemble members (20% more than the previous two embodiments to better capture strong equatorial variability signals), a horizontal localization radius of 200 km (the Rossby deformation radius is large in the equatorial region), a vertical localization radius of 250 m, and an assimilation period of 5 days (high-frequency fluctuations in the equatorial region require more frequent assimilation updates).

[0071] The assimilation strategy employs a multi-source data hierarchical weighting scheme: RAMA buoy field observations (weight 1.0, highest priority, providing direct air-sea flux and upper ocean response constraints), Argo buoy temperature and salinity profiles (weight 0.7, providing mid-ocean thermal structure), satellite altimeters (weight 0.6, providing large-scale geostrophic constraints), and SODA reanalysis (weight 0.4, providing background field coordination). In addition to 3D temperature, salinity, and current, sea surface height anomalies (SLA) are specifically added as independent assimilation variables to better constrain equatorial wave modes. The final output is a 28-year daily 3D temperature, salinity, and current assimilation data field covering 1993–2020.

[0072] In step S2, a multi-platform observation area dynamic environment database covering the collaborative network of moorings, buoys, and underwater gliders is constructed. The daily three-dimensional ocean current velocity time series for each point over 28 years is extracted from the assimilation data field. The multi-scale dynamic signal separation adopts an adaptive bandpass filtering strategy, and special frequency bands are set for special dynamic processes in the equatorial region: (1) Equatorial Kelvin wave: filtering frequency band 20-40 days, corresponding to the propagation period of the first baroclinic mode Kelvin wave across the Indian Ocean (about 25-35 days); (2) MJO (Madden-Julian oscillation): filtering frequency band 30-90 days; (3) Equatorial Rossby wave: filtering frequency band 40-120 days; (4) Near-inertial oscillation: filtering frequency band 1.5-5 days (the theoretical value of the equatorial inertial period is infinite, and the effective frequency band is actually 3-5 days); (5) High-frequency internal wave: filtering frequency band 0.3-1.5 days. In addition, annual and semi-annual cycle signals are extracted to characterize the monsoon driving features.

[0073] Vertical modal EOF decomposition was performed on the velocity fields of each frequency band (the first 5 modes were taken in the buoy region, with a cumulative variance >90%; the first 3 modes were taken in the buoy region, with a cumulative variance >85%) to obtain the time series of modal principal components. The annual maximum velocity was fitted using a Gumbel extreme value type I distribution. Since the extreme velocity distribution in the equatorial region exhibits a slight thick tail, the generalized extreme value distribution (GEV) was used as a comparison model. Model selection was performed using the AIC criterion (in this embodiment, Gumbel AIC = 156.3, GEV AIC = 153.7, the difference is not significant, and the Gumbel distribution is still used to ensure consistency with the system's standard procedure). Gumbel fitting parameters: position parameter μ = 0.1024 m / s, scale parameter σ = 0.0286 m / s, KS test D = 0.112, p = 0.698 (passed the test).

[0074] Extreme current velocities with different return periods were calculated, generating 12 typical scenarios (four more equatorial-specific scenarios than the previous two examples): climatological average, extreme Kelvin wave, MJO strong event, equatorial Rossby wave, extreme near-inertial oscillation, strong internal wave, 5-year return period extreme sea state, 10-year return period extreme sea state, 25-year return period extreme sea state, 50-year return period extreme sea state, 75-year return period extreme sea state, and 100-year return period extreme sea state. The maximum current velocity ranged from 0.0052 m / s (climatological average) to 0.3857 m / s (100-year return period extreme sea state, including the extreme case of Kelvin wave and MJO co-phase superposition). Vertical current velocity distribution for each scenario is shown.

[0075] 2. Steps S3-S4: Multi-scale spatial environment and force mapping, and time-domain solution of dynamics for multi-platform observation equipment: In step S3, parameterized finite element models are established for the three platforms, and IDW three-dimensional spatial interpolation is used to map the assimilated flow field to the nodes of each platform. The mooring model contains 286 nodes (similar to Example 1, but the vertical node distribution density is adjusted for the equatorial strong shear environment, and the node spacing within the surface 200 m is reduced to 2 m), the buoy model contains 134 nodes, and the glider model contains 78 nodes. IDW interpolation parameters: mooring and buoy N=8, p=2.0; glider, due to its large range of motion (cruising radius 15 km), adopts a dynamic interpolation strategy—updating the search range of neighboring grid points according to the glider's real-time position, N=12, p=2.5, to ensure the continuity of load mapping during the motion process.

[0076] In step S4, a unified parameterized dynamic framework for the three platforms is constructed, and synchronous simulation across multiple platforms is achieved.

[0077] (1) Deep-sea mooring platform: Total length 800 m (before optimization), main cable Φ18 mm stainless steel wire rope, equipped with 4 observation equipment cabins (located at depths of 150 m, 350 m, 550 m, and 750 m), buoyancy system with 3 24-inch glass buoys. A 6-DOF rigid body-large deformation discrete beam coupled model is adopted, and the cable is discretized into 120 Timoshenko beam elements. Morrison equation parameters: ρ_w = 1026 kg / m³ (equatorial high temperature and low salinity zone), C_D = 0.95 (cable), C_M = 2.0. Newmark-β integral, Δt = 0.04 s, total duration 7200 s.

[0078] (2) Surface buoy platform: a 3.0 m diameter disc-shaped float with a draft of 0.9 m and a displacement of 5.8 tons. The mooring system is a two-section design (upper section: 40 m polyester cable Φ22 mm; lower section: 60 m steel wire rope Φ16 mm; middle section connected by a swivel universal joint to prevent cable entanglement). The 6-DOF model of the float considers the special hydrodynamic characteristics of the low latitude region of the equator: due to the Coriolis parameter f approaching zero, the inertial-gravity wave (Poincaré wave) effect in the horizontal plane is weakened, and the coupling between vertical and horizontal motion is enhanced. Therefore, the vertical-horizontal coupling stiffness term needs to be retained in the dynamic model. The drag coefficient adopts the measured calibration value in the equatorial sea area: C_D=0.85 (disc-shaped float), C_M=1.6.

[0079] (3) Underwater glider platform: wingspan 1.5 m, fuselage length 2.0 m, net buoyancy adjustment range ±200 g. The dynamic model adopts a 6-DOF rigid body + added mass model, considering the glider's unique sawtooth profile motion mode (dive-glide-surface cycle of about 4-6 hours, typical profile depth 0-1000 m). The hydrodynamic coefficients are calibrated by CFD numerically: lift coefficient C_L=0.35 (angle of attack 15°), drag coefficient C_D=0.12, pitch moment coefficient C_Mtheta=0.08. Since the glider has active motion capability, it is regarded as a controlled rigid body in the dynamic simulation - the preset profile motion trajectory is achieved by adjusting the net buoyancy and center of gravity position, and the disturbance deviation of the extreme flow field on the preset trajectory is evaluated.

[0080] Multi-platform synchronous simulation employs a distributed coupling strategy: the three platforms independently solve their respective time-domain dynamic equations, achieving indirect coupling through the spatial correlation of the ocean current field (i.e., the dynamic response of each platform disturbs the flow field, and the disturbance is transmitted to other platforms through the continuity of the flow field). The simulation time step is uniformly set to Δt = 0.02 s, with a total duration of 10800 s (covering a complete glider profile cycle), and an output resolution of 1 s.

[0081] Dynamic response analysis of the initial design scheme under the superposition of a 0.12 m / s mesoscale eddy and MJO conditions showed that: the mooring tip drifted 112 m (exceeding the ±80 m constraint), with a maximum cable tension ratio of 0.89 (safety factor of only 1.12); the buoy drifted horizontally by 68 m, with a mooring angle of 35°, exceeding the sensor stability requirements; the glider's preset profile trajectory deviated by 42 m (horizontal direction), causing it to fail to traverse the mooring observation area along the predetermined path, resulting in a collaborative observation spatial coverage of only 65.4% (target value >90%). The system determined that the initial scheme did not meet the requirements for multi-platform collaborative observation, triggering joint optimization.

[0082] 3. Step S5, multibody dynamics closed-loop optimization iteration: The multibody dynamics closed-loop optimization iteration module designs an extended dual-objective optimization framework for multi-platform collaborative scenarios: the primary objective is the collaborative observation space coverage f_coverage (defined as the ratio of the union of the effective observation volumes of the three platforms to the total volume of the target area, with a target value > 90%), and the secondary objective is the weighted sum of the structural safety indicators f_safety of each platform (0.4 for underwater moorings, 0.35 for buoys, and 0.25 for gliders). The NSGA-III algorithm is used (compared to NSGA-II, it is more suitable for high-dimensional optimization problems with three or more objectives; this embodiment is essentially a dual objective of coverage + safety, but involves the decomposition of sub-objectives for three platforms). The population size is NP = 120, the maximum number of iterations is G_max = 250, the crossover probability is P_c = 0.88 (SBX η_c = 18), and the mutation probability is P_m = 0.10 (polynomial mutation η_m = 18).

[0083] The equipment key design parameter set Θ contains three sets of platform parameters with a total of 15 optimization variables. The submersible mooring group (6 variables): total length L_s (500-900 m), main cable diameter D_s (Φ14-Φ22 mm), number of buoys N_s (3-8), buoy specifications S_s (17 / 24 / 28 inches), equipment compartment spacing Δz_s (20-80 m), and anchor weight W_s (200-800 kg). The buoy group (5 variables): diameter D_f (2.6-3.5 m), hull shape T_f (disc / hemispherical / cylindrical), mooring method M_f (single / two / three sections), mooring length L_f (80-140 m), and anchor weight W_f (300-600 kg). Glider group (4 variables): cruising depth range Z_g (100-1500 m), net buoyancy adjustment range B_g (±100-±300 g), wingspan-to-aspect ratio AR_g (4-8), and profile glide angle θ_g (15-30°).

[0084] The optimization process is executed in three phases: Phase 1 (generations 1-80) is platform-level independent optimization, where each platform optimizes its own parameters to meet single-platform safety constraints; Phase 2 (generations 81-180) is collaborative coupling optimization, fixing the initial optimization parameters of each platform and focusing on adjusting the spatial geometry configuration between platforms (deployment spacing, depth layering strategy) to improve coverage; Phase 3 (generations 181-250) is global fine-tuning convergence, where the parameters of each platform are finely adjusted within the collaborative framework to pursue a uniform distribution of the Pareto front. Each generation evaluation requires 36 complete dynamic simulations across three platforms × 12 scenarios, resulting in a large computational load. The system employs parallel computing for acceleration (independent simulations of the three platforms are executed in parallel).

[0085] After 198 iterations, the system determined that the Pareto front converged (Hypervolume change rate <0.03% for 20 consecutive generations). The final compromise optimal solution was selected, with the following optimization scheme: the total length of the mooring buoy was optimized to 620 m (a reduction of 22.5%), the main cable was upgraded to a Φ20 mm high-strength steel wire rope, and five 28-inch glass buoys were configured (two in parallel at the top, one at 250 m depth, and two at 450 m depth), with a total net buoyancy of 466.4 kg; an 800 kg gravity anchor at the bottom was used in conjunction with dual acoustic release devices. The buoy was changed to a disc-shaped design with a diameter of 2.8 m, and the mooring was upgraded to a three-section polyester cable (total length 90 m) with a 450 kg anchor weight. The glider's cruising depth was adjusted to 200-1000 m (avoiding the surface strong shear layer), with a net buoyancy adjustment range of ±220 g, a wingspan-to-aspect ratio of 6.5, and a glide angle of 22°. The dynamic optimization results of the multi-platform collaborative observation system are shown in Table 3.

[0086] Table 3 Configuration of Dynamics Optimization Results for Multi-Platform Collaborative Observation System

[0087] 4. Closed-loop verification effect: The multi-platform collaborative optimization scheme was applied to 12 typical scenarios for full-coverage closed-loop verification. The dynamic response verification results of the optimized multi-platform collaborative system in the 12 typical scenarios are as follows: Under the climatological average scenario (maximum current velocity 0.0052 m / s), all three platforms were in excellent stability: the mooring tip drifted 2.8 m, the buoy drifted horizontally 1.5 m, the glider's profile trajectory deviated 3.2 m, and the cooperative space coverage reached 94.7%. Under the extreme Kelvin wave scenario (maximum current velocity 0.1856 m / s), the mooring tip drifted 28.6 m, the buoy drifted 15.3 m, the glider's trajectory deviated 18.5 m, and the coverage was 89.2%. Under the MJO strong event scenario (maximum current velocity 0.2234 m / s), the mooring tip drifted 35.2 m, the buoy drifted 19.8 m, the glider deviated 22.4 m, and the coverage was 87.6%.

[0088] Under extreme sea conditions occurring once every 50 years (maximum current velocity 0.3245 m / s), the responses of each platform are as follows: the mooring tip displacement was 51.3 m (constraint <80 m), the maximum cable tension ratio was 0.58 (safety factor 1.72), and the structure maintained a smooth catenary; the buoy horizontal displacement was 34.6 m, the mooring inclination angle was 16.8°, and the float inclination angle was 5.4°, meeting the sensor requirements; the glider trajectory deviated by 28.7 m, which was reduced to 19.3 m after adaptive buoyancy adjustment compensation, still effectively traversing the mooring observation area. The three platforms achieved a collaborative spatial coverage of 84.5%.

[0089] Under the most severe 100-year return period extreme sea state (maximum current velocity 0.3857 m / s, Kelvin wave and MJO phase superposition): the mooring tip displacement was 58.6 m, tension ratio 0.72 (safety factor 1.39); the buoy horizontal displacement was 42.1 m, mooring inclination angle was 21.3°, and float inclination angle was 7.8°; the glider trajectory deviation was 36.8 m, which was reduced to 26.5 m after adaptive compensation. All three platform structures remained safe with no risk of instability. The cooperative space coverage was 78.3%, which, although lower than the normal level, is still acceptable under this extremely rare operating condition.

[0090] Comparison of core performance before and after optimization: The multi-platform collaborative space coverage increased from 65.4% (normal) before optimization to 92.3% (normal) after optimization, an improvement of 41.1%; the drift of the moored buoy in normal currents decreased from 112 m to 11.3 m (improvement of 89.9%), the drift of the buoy decreased from 68 m to 6.8 m (improvement of 90.0%), and the trajectory deviation of the glider decreased from 42 m to 12.6 m (improvement of 70.0%). All three platforms met their respective structural safety constraints in all 12 typical scenarios, achieving efficient, safe, and stable operation of the multi-platform collaborative observation system in the complex dynamic environment of the Indian Ocean equator.

[0091] The technical features of the examples described above can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the examples are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The examples described above merely illustrate embodiments of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and non-substantial improvements without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A dynamic system for multi-platform observation equipment based on marine environmental assimilation data, characterized in that, It includes a marine environment assimilation data generation module, a marine dynamic feature extraction module for equipment deployment area, a multi-scale spatial environment and force mapping module, a time-domain solution module for the dynamics of multi-platform observation equipment, and a multi-body dynamics closed-loop optimization iteration module; the multi-body dynamics closed-loop optimization iteration module feeds back the optimization data to the multi-scale spatial environment and force mapping module. The marine environment assimilation data generation module acquires multi-source heterogeneous data within the observation area and extracts the air-sea boundary forcing conditions. Based on the regional ocean circulation numerical model, it constructs a three-dimensional grid model of the target sea area and outputs a long-sequence standardized marine environment analysis field dataset with dynamic consistency. The marine dynamic feature extraction module for the equipment deployment area receives the historical three-dimensional ocean current velocity time series of the target sea area from the standardized marine environment analysis field dataset, separates the marine dynamic signals of different frequency bands through multi-scale bandpass filtering, and analyzes the abnormal signals of the current velocity field to generate a three-dimensional flow field model covering multiple working conditions of extreme dynamic scenarios. The multi-scale spatial environment and force mapping module uses a three-dimensional spatial interpolation method to map the macroscopic grid-scale flow velocity to the microscopic finite element nodes of the parameterized equipment model and calculate the node inflow velocity vector. The aforementioned multi-platform observation equipment dynamics time-domain solution module constructs a unified parameterized dynamics simulation framework compatible with multiple types of observation equipment, solves fluid-structure interaction nonlinear fluid forces and multi-physics field coupled loads, constructs a time-domain dynamic model of a rigid body-discrete beam coupled system, and outputs the equipment's full-time three-dimensional displacement field and cable tension field; The multibody dynamics closed-loop optimization iteration module constructs a dual-objective cost evaluation function that includes a structural safety evaluation function and an observation stability evaluation function. It then uses a multi-objective genetic algorithm to iteratively optimize the key design parameter set until the Pareto front converges, outputting the optimal equipment structural configuration parameter scheme.

2. A dynamic method for multi-platform observation equipment based on marine environmental assimilation data, characterized in that, The dynamic system of multi-platform observation equipment based on marine environmental assimilation data as described in claim 1 performs the following steps: Step S1: Acquire multi-source heterogeneous data within the observation area and extract the air-sea boundary forcing conditions. Construct a three-dimensional grid model of the target sea area based on the regional ocean circulation numerical model. Assimilate and fuse multi-element ocean observation data to output a long-sequence standardized ocean environment analysis field dataset with dynamic consistency. Step S2: Receive the sea area coordinate parameters of the observation equipment deployment point, extract the historical three-dimensional ocean current velocity time series of the target sea area from the standardized marine environment analysis field dataset output in Step S1, separate the ocean dynamic signals of different frequency bands through multi-scale bandpass filtering, analyze the current velocity field anomaly signal for vertical spatial projection, and then combine the extreme value statistical distribution to fit the historical maximum current velocity, calculate the theoretical extreme current velocity amplitude of different return periods, and generate a three-dimensional flow field model covering multiple working conditions of extreme dynamic scenarios; Step S3: Using a three-dimensional spatial interpolation method, the macroscopic mesh-scale velocity in the three-dimensional flow field model of the extreme dynamic scenario generated in step S2 is mapped to the microscopic finite element nodes of the parameterized equipment model, and the inflow velocity vector at each node is calculated. Step S4: Construct a unified parameterized dynamic simulation framework compatible with multiple types of observation equipment. Using the core hydrodynamic stability parameters of the observation equipment mapped in Step S3 as load inputs, solve for fluid-structure interaction nonlinear fluid forces and multi-physics coupling loads. Abstract the buoyancy components and equipment compartment of the equipment as rigid body elements and the cables as discrete beam elements. Combine the above multi-physics coupling loads to construct and solve the time-domain dynamic model of the rigid body-discrete beam coupled system, and output the full-time three-dimensional displacement field of the equipment and the cable tension field. Step S5: Construct a dual-objective cost evaluation function that combines the structural safety evaluation function and the observation stability evaluation function. Using a multi-objective genetic algorithm, substitute the full-time three-dimensional displacement field and cable tension field output in Step S4 into the cost function for evaluation. Set safety constraints, iteratively optimize the set of key equipment design parameters, generate a new parameter set, and automatically feed it back to Step S3 for closed-loop recalculation until the Pareto front converges, and output the optimal equipment structural configuration parameter scheme.

3. The dynamic method for multi-platform observation equipment based on marine environmental assimilation data according to claim 2, characterized in that, The multi-source heterogeneous data in step S1 includes satellite remote sensing, field measurements, and reanalysis products; multiple elements include temperature, salinity, and flow velocity; and long sequences with dynamic consistency include three-dimensional flow velocity vector fields, temperature and salinity fields, and sea surface wind fields.

4. The dynamic method for multi-platform observation equipment based on marine environmental assimilation data according to claim 2, characterized in that, The historical three-dimensional ocean current velocity time series in step S2 is the three-dimensional ocean current velocity time series of the sea area where the deployment point is located; the ocean dynamic signal is the ocean dynamic signal of different frequency bands separated by multi-scale bandpass filtering; the three-dimensional flow field model of extreme dynamic scenarios includes a multi-scenario three-dimensional flow field model covering normal to extreme sea states.

5. The dynamic method for multi-platform observation equipment based on marine environmental assimilation data according to claim 2, characterized in that, The observation equipment in step S4 includes deep-sea moorings, observation buoys, and marine unmanned aerial vehicles (UAVs); the core parameters of the hydrodynamic stability of the observation equipment include nodal inflow velocity, sea surface wind field, equipment weight and buoyancy configuration; the multi-physics coupling loads include sea surface wind load, gravity and buoyancy configuration.

6. The dynamic method for multi-platform observation equipment based on marine environmental assimilation data according to claim 2, characterized in that, The set of key equipment design parameters in step S5 includes equipment geometry, buoyancy configuration, cable segmentation, and anchorage structure; the safety assessment function is constructed by weighting the equipment structural strength and attitude stability. The observation stability evaluation function is a dynamic displacement measure of key detection nodes; the safety constraint condition is that all cables meet the safety factor threshold constraint.

Images