Underwater vehicle oriented complex domain hybrid density grid flow field modeling method

Abstract

The application discloses a complex domain hybrid density grid flow field modeling method for underwater vehicles, comprising the following steps: acquiring terrain boundary, water depth, incoming flow, free surface and complex water area constraint region data of a water area to be modeled, and establishing a three-dimensional numerical model; dividing sub-computing domains according to the complex water area constraint region, the free surface, the near-wall boundary layer and the main flow region, generating locally encrypted grids for the complex boundary, the free surface and the near-wall region, and generating main flow region grids for the main flow region; interpolating and transferring flow velocity, pressure and phase volume fraction between adjacent interface boundaries of different grid density sub-computing domains to form a partitioned variable density grid flow field calculation model; performing unsteady solving by using a RANS control equation model; and extracting a target vehicle depth layer horizontal flow velocity component to generate flow field query results for path planning or track tracking control of the underwater vehicle.

Description

Technical Field

[0001] This invention relates to the field of underwater vehicle environment modeling and flow field prediction technology, specifically to a method for modeling complex key domain mixed density grid flow fields for underwater vehicles. Background Technology

[0002] Autonomous underwater vehicles (AUVs) require information on flow velocity, flow direction, free surface, and the influence of complex boundaries within the mission area for applications such as near-shore inspection, port navigation, channel passage, missions around islands and reefs, and exploration of complex waters. This flow field information can serve as environmental input for path planning, trajectory tracking and control, energy consumption estimation, and risk assessment of the mission area.

[0003] Complex waterways typically contain constrained areas such as port entrances and exits, wharves, channels, islands, reefs, coastlines, near-shore structures, and areas of abrupt changes in underwater topography. These areas can lead to phenomena such as boundary layer separation, flow around the flow, backflow, local vortex structures, free surface disturbances, and abrupt changes in water depth, resulting in a flow field that exhibits significant three-dimensionality, unsteadiness, and spatial inhomogeneity.

[0004] In existing flow field modeling, computational fluid dynamics methods are typically used to establish the computational domain and mesh it, then the flow field results are obtained by solving the governing equations. However, when the computational domain contains both large-scale main flow regions and local complex boundary regions, if a uniform high-density mesh is used throughout the domain, the computational scale is large; if a low-density mesh is used throughout the domain, the flow field characteristics at complex boundaries, free surfaces, or near-wall boundary layers are difficult to fully analyze.

[0005] Furthermore, traditional linear wave or simplified flow field prediction methods are insufficient to fully describe complex three-dimensional flows below the free surface, and they also struggle to balance computational efficiency for water-air two-phase free surfaces, near-wall turbulence, complex boundary flows, and the main flow region. For applications targeting underwater vehicles, obtaining only three-dimensional flow field calculation results is not enough; it is also necessary to convert the three-dimensional time-varying velocity field into queryable and accessible flow field data at the target navigation depth.

[0006] Therefore, a flow field modeling method is needed that can perform partitioned modeling for complex water area constraints, generate different grid densities according to regional characteristics, perform variable interpolation and transfer at the interface of different grid density sub-computation domains, and output the flow field query results of the target navigation depth layer. This method can reduce the overall computational load while ensuring the flow field analysis capability of key areas, and provide callable flow field information for underwater vehicle path planning or trajectory tracking control. Summary of the Invention

[0007] The purpose of this invention is to provide a method for modeling complex key domain flow fields using mixed density grids for underwater vehicles. This method aims to solve the problems in existing complex water flow field modeling, such as the large computational scale of uniform high-density grids, the difficulty of fully resolving flow field characteristics at complex boundaries and free surfaces using low-density grids, and the lack of reasonable interface variable transfer mechanisms between different density grid regions, resulting in insufficient variable connection.

[0008] To achieve the above objectives, the present invention employs the following technical solution: A method for modeling complex key-area mixed-density grid flow fields for underwater vehicles, including: The environmental data of the water body to be modeled is obtained, including topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and data of complex water body constraint areas; A three-dimensional numerical model of the water area to be modeled is established based on the environmental data, and the three-dimensional numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary and complex water area constraints. Based on the geometric boundaries, free surface location, near-wall boundary layer location, and main flow region location of the complex water area constraint region, the computational domain corresponding to the three-dimensional numerical model is divided into multiple sub-computational domains, and adjacent sub-computational domains are coupled through interface boundaries. A localized mesh with a first mesh density is generated for a sub-computational domain containing the geometric boundary of the complex water area constraint region, the free liquid surface, or the near-wall boundary layer. A main flow region mesh with a second mesh density is generated for a sub-computational domain that is far from the geometric boundary of the complex water area constraint region and located in the main flow region. The first mesh density is greater than the second mesh density. Between adjacent computational domains with different grid densities, flow field variables are interpolated and transferred through interface boundaries, and phase volume fractions are interpolated and transferred when the interface boundaries involve free liquid surfaces, so as to form a partitioned variable density grid flow field computational model. Based on the partitioned variable density grid flow field calculation model, RANS governing equations and SST are used. - The unsteady flow field was solved using the two-equation turbulence model and the VOF multiphase flow model, resulting in the three-dimensional time-varying velocity field of the water body to be modeled. Based on the target navigation depth layer of the underwater vehicle, flow field data corresponding to the depth layer is extracted from the three-dimensional time-varying velocity field, and spatial interpolation is performed on the flow field data to generate flow field query results for underwater vehicle path planning or trajectory tracking control.

[0009] Furthermore, the sub-computational domain includes a complex boundary sub-computational domain, a free surface sub-computational domain, a near-wall boundary layer sub-computational domain, and a main flow sub-computational domain. The complex boundary sub-computational domain is used to cover local areas near port entrances and exits, wharves, waterways, islands, reefs, coastlines, port boundaries, near-shore structures, or areas with abrupt changes in underwater topography. The free surface sub-computational domain is used to cover areas near the water-air two-phase interface. The near-wall boundary layer sub-computational domain is used to cover areas near the non-slip boundary of the wall. The main flow sub-computational domain is used to cover areas far from the complex boundary and where the flow changes gently. The division of the sub-computational domain is based on the distance from the geometric boundary of the complex water area constraint region to the spatial location within the computational domain, the curvature of the geometric boundary, the location of the free surface, the normal distance of the near-wall region, and the range of the main flow region.

[0010] Furthermore, when the distance from a spatial location within the computational domain to the geometric boundary of the complex water area constraint region is less than a first distance threshold, the spatial location is assigned to a localized denser region corresponding to the first grid density; when the distance is greater than or equal to the first distance threshold and less than the second distance threshold, the spatial location is assigned to a transitional grid density region; when the distance is greater than or equal to the second distance threshold and is located in the main flow region, the spatial location is assigned to the main flow region corresponding to the second grid density; wherein, the first distance threshold is less than the second distance threshold, and the first and second distance thresholds are set according to the water area scale, boundary curvature, expected flow field gradient, and computational resources.

[0011] Furthermore, the localized mesh includes a localized mesh generated near the free surface, a localized mesh generated at the geometric boundaries corresponding to complex shorelines, islands, reefs, docks, or waterways, and a boundary layer mesh generated near a non-slip wall. The localized mesh is generated using one or more of the following methods: local size constraints, local influence volumes, boundary layer meshes, curvature control meshes, or proximity control meshes. The mesh of the main flow region is generated using unstructured meshes, polyhedral meshes, hexahedral meshes, or combinations thereof.

[0012] Furthermore, the interpolation transfer of flow field variables through the interface boundary includes: for velocity variables, interpolation mapping is performed based on the positional relationship between the centers of the grid surfaces or grid cells on both sides of the interface, or area-weighted interpolation is performed based on the overlapping area of ​​the interface; for pressure variables, interpolation is performed based on the pressure values ​​of the adjacent cells of the interface, and the interpolation results are used for pressure-velocity coupling solutions; for phase volume fractions, interpolation is performed at the interface boundary involving the free liquid surface, and the interpolated phase volume fractions are subject to range constraints; when the interpolation result is less than 0, it is truncated to 0, and when the interpolation result is greater than 1, it is truncated to 1, or the phase volume fractions in the neighborhood of the interface are normalized so that the volume fractions of each phase satisfy the constraint that they add up to 1; the interface boundary is an interface boundary with matched grid nodes or a non-matched interface boundary with incompletely matched grid nodes.

[0013] Furthermore, the SST - Two-equation turbulence model with turbulent kinetic energy Sum of dissipation To solve for variables, the model is used to simulate flow separation, backflow, and wall turbulence characteristics in locally refined grids and near-wall boundary layer refined grids within the complex water-constrained region. The VOF multiphase flow model is used to track the water-air two-phase flow free surface in the locally refined grid corresponding to the free surface, when calculating the phase volume fraction in the grid. satisfy When the calculation grid is defined as a free liquid surface interface grid, the mixing density is determined by weighting the density and volume fraction of each phase fluid for the water-air two-phase flow interface grid.

[0014] Furthermore, the boundary conditions for solving the unsteady flow field include inlet velocity boundary conditions, outlet pressure boundary conditions, wall no-slip boundary conditions, and symmetric conditions for the outer boundary of the air domain when an air domain exists. The unsteady flow field solution uses the finite volume method to discretize the flow control equations and uses a pressure-velocity coupling algorithm to solve the pressure and velocity fields. When the partitioned variable density grid flow field calculation model contains grids with large aspect ratios, grids with large volume differences between adjacent elements, or polyhedral mixed grids, a gradient correction method is used to solve the gradient of the flow field variables.

[0015] Furthermore, the flow field query results are obtained by querying the flow field query table; the flow field query table is built based on location coordinates and time index, and is used to output the predicted flow velocity value at the query location; The query location includes underwater vehicle planned path nodes, candidate path nodes, or mission area sampling nodes; when the query location is not located at an existing sampling point or grid node, spatial interpolation is performed based on the velocity components of the sampling points or grid nodes around the query location; when the query time is not equal to the existing calculation time, temporal interpolation is performed based on the flow field data of adjacent calculation times; the spatial interpolation adopts at least one of cubic spline interpolation, bilinear interpolation, bicubic interpolation, or radial basis function interpolation.

[0016] A complex water area partitioned variable density grid flow field modeling system for underwater vehicles, based on the aforementioned method, includes an environmental data acquisition module, a 3D model construction module, a sub-computation domain partitioning module, a variable density grid generation module, an interface variable transfer module, an unsteady flow field solution module, and a flow field result output module. The environmental data acquisition module is used to acquire environmental data of the water area to be modeled. The environmental data includes topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and complex water area constraint region data. The three-dimensional model building module is used to establish a three-dimensional numerical model of the water area to be modeled based on the environmental data. The three-dimensional numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary, and complex water area constraints. The sub-computation domain partitioning module is used to divide the computation domain corresponding to the three-dimensional numerical model into multiple sub-computation domains according to the geometric boundary of the complex water area constraint region, the location of the free liquid surface, the location of the near-wall boundary layer, and the location of the main flow region, and to couple adjacent sub-computation domains through interface boundaries. The variable density mesh generation module is used to generate a locally refined mesh with a first mesh density for a sub-computation domain containing the geometric boundaries of complex water area constraints, free liquid surfaces, or near-wall boundary layers, and to generate a main flow region mesh with a second mesh density for a sub-computation domain containing the main flow region, wherein the first mesh density is greater than the second mesh density. The interface variable transfer module is used to interpolate and transfer flow field variables, including at least flow velocity and pressure, between adjacent sub-computation domains with different grid densities through interface boundaries, and to interpolate and transfer phase volume fraction when the interface boundary involves a free liquid surface. The unsteady flow field solution module is used to solve RANS governing equations and SST. - The unsteady flow field was solved by using the two-equation turbulence model and the VOF multiphase flow model to obtain the three-dimensional time-varying velocity field. The flow field result output module is used to extract flow field data including at least the horizontal velocity component based on the target navigation depth layer of the underwater vehicle, and generate flow field query results for underwater vehicle path planning or trajectory tracking control through spatial interpolation.

[0017] An electronic device includes a processor, a memory, and a computer program stored in the memory; when the processor executes the computer program, it implements the complex domain hybrid density grid flow field modeling method for underwater vehicles.

[0018] A computer-readable storage medium storing a computer program; when executed by a processor, the computer program implements the complex key domain hybrid density grid flow field modeling method for underwater vehicles.

[0019] Compared with the prior art, the present invention has the following technical features: 1. This invention divides the computational domain into sub-domains according to the complex water area constraint region, free liquid surface, near-wall boundary layer and main flow region, and sets a first grid density and a second grid density. This invention can improve the grid resolution capability in complex boundary, free liquid surface and near-wall regions, while avoiding the use of unnecessary global high-density grids in the main flow region.

[0020] 2. This invention interpolates and transfers velocity, pressure, and phase volume fraction between the interface boundaries of adjacent computational domains with different grid densities. This invention enables the velocity field, pressure variables, and free surface interface positions to connect with each other in different grid density regions, which is beneficial to improving the solution continuity of the partitioned variable density grid flow field computational model.

[0021] 3. By combining the RANS governing equations, the SSTk-ω two-equation turbulence model, and the VOF multiphase flow model on the partitioned variable density grid flow field calculation model, this invention can take into account unsteady flow, near-wall turbulence, and water-air free surface tracking in complex water areas. It is applicable to complex water scenarios such as ports, waterways, islands, reefs, and nearshore structures.

[0022] 4. By extracting the flow field data of the target navigation depth layer from the three-dimensional time-varying flow velocity field and establishing a flow field lookup table based on position coordinates and time index, this invention can convert the three-dimensional flow field solution results into a data form that can be called by underwater vehicle path planning or track tracking control. Attached Figure Description

[0023] Figure 1 This is a schematic flowchart of the method of the present invention; Figure 2 This is a schematic diagram of the complex water area partitioned variable density grid flow field modeling system in an embodiment of the present invention; Figure 3 This is a schematic diagram of the sub-computation domain partitioning and interface boundary coupling of complex water areas in an embodiment of the present invention; Figure 4 This is a schematic diagram of the first grid density locally refined grid and the second grid density main flow region grid in an embodiment of the present invention; Figure 5 This is a schematic diagram illustrating the process of generating the target navigation depth layer flow field lookup table in an embodiment of the present invention. Detailed Implementation

[0024] Existing methods for modeling complex water flow fields struggle to simultaneously balance critical flow field accuracy, overall computational scale, and target depth layer query output required for underwater vehicle applications when dealing with both large-scale main flow regions and locally complex boundary regions. To address this issue, this invention provides a method for modeling complex key regions using a hybrid density grid. The overall process is as follows: First, acquire environmental data of the water area to be modeled, including topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and complex water constraint region data. Second, establish a three-dimensional numerical model based on the environmental data, including water topography, inlet boundary, outlet boundary, free surface, wall boundary, and complex water constraint region. Third, divide the computational domain into multiple sub-computational domains based on the geometric boundaries of the complex water constraint region, the location of the free surface, the location of the near-wall boundary layer, and the location of the main flow region, coupling adjacent sub-computational domains through interface boundaries. Fourth, model the complex water... The method generates locally refined grids with a first grid density for sub-computational domains at the geometric boundaries of the constrained region, the free surface, or the near-wall boundary layer. For the sub-computational domains of the main flow region, a main flow region grid with a second grid density is generated. Velocity, pressure, and phase volume fraction are interpolated and transferred between adjacent sub-computational domains with different grid densities through interface boundaries, forming a partitioned variable-density grid flow field calculation model. Unsteady flow fields are solved using the RANS governing equations, the SSTk-ω two-equation turbulence model, and the VOF multiphase flow model to obtain a three-dimensional time-varying velocity field. Horizontal velocity components are extracted based on the target navigation depth layer of the underwater vehicle, and spatial interpolation is used to generate flow field query results for path planning or trajectory tracking control. This method balances local flow field analytical capabilities with overall computational scale.

[0025] like Figure 1 As shown, the method provided in this embodiment includes the following steps.

[0026] S1, acquire environmental data.

[0027] The environmental data of the water body to be modeled is obtained, including topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and data of complex water body constraint areas.

[0028] Specifically, topographic boundary data is used to determine the spatial boundaries of shorelines, islands, reefs, port boundaries, or nearshore structures; water depth data is used to determine the topography of the seabed or lakebed; inflow velocity data and inflow direction data are used to determine the inlet boundary conditions; and free surface location data is used to determine the initial location of the water-air two-phase interface or local densification areas.

[0029] The complex water area constraints include at least one of the following: port entrances / exits, wharves, channels, islands, reefs, coastlines, port boundaries, nearshore structures, and areas of abrupt changes in underwater topography; the nearshore structures include the boundary structures corresponding to ports, wharves, or coastlines. These areas typically exhibit complex geometric boundaries, significant local curvature changes, strong near-wall effects, and pronounced flow separation or backflow phenomena, thus serving as a basis for subsequent sub-computation domain partitioning and local densification.

[0030] S2, Establish a three-dimensional numerical model.

[0031] Based on the environmental data obtained in step S1, a three-dimensional numerical model of the water area to be modeled is established, such that the three-dimensional numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary, and complex water area constraints.

[0032] Specifically, the shoreline shape, seabed or lakebed topography, and water boundaries of the water area to be modeled are determined based on the topographic boundary data and water depth data; the inlet boundary is determined based on the incoming flow velocity data and incoming flow direction data; the outlet boundary is determined based on the extension direction of the computational domain or the outflow direction; and the corresponding geometric region is formed in the three-dimensional numerical model based on the complex water area constraint region data.

[0033] S3, partition the computational domain.

[0034] Based on the geometric boundaries, free surface location, near-wall boundary layer location, and main flow region location of the complex water area constraint region obtained in step S1, the computational domain corresponding to the three-dimensional numerical model established in step S2 is divided into multiple sub-computational domains, and adjacent sub-computational domains are coupled through interface boundaries.

[0035] Specifically, the sub-computational domains include a complex boundary sub-computational domain, a free surface sub-computational domain, a near-wall boundary layer sub-computational domain, and a main flow sub-computational domain. This allows the complex boundary, free surface, near-wall boundary layer, and main flow regions to be meshed according to their respective flow field variation characteristics. The complex boundary sub-computational domain is used to cover local areas near port entrances / exits, wharves, channels, islands, reefs, coastlines, port boundaries, near-shore structures, or areas with abrupt changes in underwater topography. The free surface sub-computational domain is used to cover areas near the water-air two-phase interface. The near-wall boundary layer sub-computational domain is used to cover areas near the non-slip boundary of the wall. The main flow sub-computational domain is used to cover areas far from the complex boundary where the flow changes are relatively gentle.

[0036] like Figure 3 As shown, the complex boundary sub-computational domain, the free surface sub-computational domain, the near-wall boundary layer sub-computational domain, and the main flow sub-computational domain are connected by interface boundaries. These interface boundaries are used to transfer flow field variables between adjacent sub-computational domains and maintain the overall coupling of the computational domains. The sub-computational domains can be geometric partitions, mesh generation partitions, or solution partitions in a three-dimensional numerical model. They do not require complete physical isolation; as long as different mesh densities can be set according to regional characteristics and variables can be connected through interface boundaries, it is acceptable.

[0037] The division is based on the following criteria: Sub-computational domains are determined according to the distance from the geometric boundary of the complex water area constraint region to its spatial location within the computational domain, the curvature of the geometric boundary, the location of the free surface, the normal distance to the near-wall region, and the range of the main flow region. These include determining the complex boundary sub-computational domain, the free surface sub-computational domain, the near-wall boundary layer sub-computational domain, and the main flow sub-computational domain. The mesh density level of the corresponding sub-computational domain is determined by comparing the distance, geometric boundary curvature, or near-wall region normal distance with a preset threshold. The preset thresholds for the distance, geometric boundary curvature, free surface neighborhood range, or near-wall region normal distance can be determined based on the water area scale, modeling accuracy requirements, computational resource constraints, and the importance of the task area.

[0038] In one implementation, when the distance from a spatial location within the computational domain to the geometric boundary of a complex water area constraint region is less than the corresponding boundary distance threshold, the spatial location is assigned to a complex boundary sub-computational domain; when the curvature of the geometric boundary is greater than the corresponding curvature threshold, the mesh density level of the sub-computational domain at that location is increased; when the spatial location is within the neighborhood of the free surface, it is assigned to the free surface sub-computational domain; when the normal distance from the spatial location to the wall boundary is less than the near-wall normal distance threshold, it is assigned to the near-wall boundary layer sub-computational domain.

[0039] In one embodiment, if the distance from a spatial location to the geometric boundary of a complex water area constraint region is less than a first distance threshold, the spatial location is assigned to a locally refined region corresponding to a first grid density; if the distance is greater than or equal to the first distance threshold and less than a second distance threshold, the spatial location is assigned to a transitional grid density region; if the distance is greater than or equal to the second distance threshold and is located within the main flow region, the spatial location is assigned to the main flow region corresponding to a second grid density. The first distance threshold is less than the second distance threshold. The first and second distance thresholds can be set based on the water area scale, boundary curvature, expected flow field gradient, and computational resources.

[0040] S4 generates a partitioned variable density mesh.

[0041] A localized mesh with a first mesh density is generated for a sub-computational domain containing the geometric boundary of the complex water area constraint region, the free surface, or the near-wall boundary layer in step S1. A main flow region mesh with a second mesh density is generated for a sub-computational domain that is far from the geometric boundary of the complex water area constraint region and located in the main flow region. The first mesh density is greater than the second mesh density.

[0042] Specifically, such as Figure 4 As shown, a local fine mesh is generated near the free surface, a local fine mesh is generated at the geometric boundaries corresponding to complex shorelines, islands, reefs, docks or waterways, a boundary layer fine mesh is generated near the non-slip wall, and a main flow region mesh with a density lower than the local fine mesh is generated in the main flow region far away from the complex boundary and where the flow changes relatively gently.

[0043] In one embodiment, the grid density level can be determined based on the grid scale or the number of grid cells per unit volume. If the grid scale corresponding to the first grid density is... The grid scale corresponding to the second grid density is ,but If the number of grid cells per unit volume corresponding to the first grid density is The second grid density corresponds to the number of grid cells per unit volume. ,but For the complex boundary subdomain, free surface subdomain, and near-wall boundary layer subdomain, a first grid density can be set; for the main flow subdomain, a second grid density can be set; and for the transition region between the two, one or more intermediate grid density levels can be set. These settings improve grid resolution in complex boundary, free surface, and near-wall regions, while controlling the grid size in the main flow region.

[0044] The first and second grid densities can be two discrete levels, or two adjacent or non-adjacent levels within a multi-level grid density. The locally refined grid can be generated using one or more of the following methods: local size constraints, local influence volumes, boundary layer grids, curvature control grids, or proximity control grids. The main flow region grid can be an unstructured grid, a polyhedral grid, a hexahedral grid, or a combination thereof.

[0045] S5, Interface Variable Interpolation Passing.

[0046] Between adjacent sub-computational domains with different grid densities, interpolation is performed to transfer flow field variables, including at least velocity and pressure, through interface boundaries. When the interface boundary involves a free surface, interpolation is also performed to transfer phase volume fraction, thus forming a partitioned variable-density grid flow field computational model. The interface boundaries connect adjacent sub-computational domains with different grid densities, and the interpolation transfer ensures that the velocity fields, pressure variables, and free surface interface positions are interconnected between adjacent regions with different grid densities.

[0047] Specifically, the interface boundary is the boundary between adjacent sub-computation domains used to exchange or map flow field variables. The interface boundary can be a matching interface boundary or a non-matching interface boundary where the mesh nodes are not fully matched. For adjacent sub-computation domains with different mesh densities or non-matching mesh cells, velocity, pressure, and phase volume fraction are transferred between the interface boundaries of adjacent sub-computation domains through linear interpolation, area-weighted interpolation, volume-weighted interpolation, conservative interpolation, or other interpolation methods.

[0048] The velocity and pressure used to solve the RANS governing equations are interpolated and transferred between the interface boundaries of adjacent sub-computation domains. At the interface boundaries where there is a water-air two-phase interface, the phase volume fraction used to solve the VOF multiphase flow model is interpolated and transferred, thereby connecting the velocity field, pressure variables and free surface interface positions between adjacent regions with different grid densities.

[0049] In one embodiment, for velocity variables, interpolation mapping is performed based on the positional relationship between the centers of the mesh surfaces or mesh cells on both sides of the interface, or area-weighted interpolation is performed based on the overlapping area of ​​the interface to obtain the velocity value on the target interface mesh; for pressure variables, interpolation is performed based on the pressure values ​​of the adjacent cells of the interface, and the interpolation result is used for pressure-velocity coupling solution; for phase volume fraction, interpolation is performed at the interface boundary involving the free liquid surface, and the interpolated phase volume fraction is range-constrained to ensure that it falls within the specified range. to The above method can reduce the problem of variable discontinuity caused by different grid densities or mismatched grid nodes in adjacent sub-computation domains.

[0050] In one embodiment, after interpolating the phase volume fraction using an interface, if the interpolation result is less than... Then cut it into If the interpolation result is greater than Then cut it into Alternatively, the phase volume fractions within the interface neighborhood can be normalized so that the sum of the phase volume fractions is equal to the sum of the phase volume fractions. This constraint reduces interface position jumps that occur when the free surface crosses computational domains with different mesh densities.

[0051] S6, Solving unsteady flow fields.

[0052] Based on the partitioned variable density grid flow field calculation model formed in step S5, the unsteady flow field is solved using the RANS governing equations, the SSTk-ω two-equation turbulence model, and the VOF multiphase flow model to obtain the three-dimensional time-varying velocity field of the water body to be modeled. The RANS governing equations describe the average flow of incompressible fluids, the SSTk-ω two-equation turbulence model simulates the turbulence characteristics in complex water confined regions and near-wall boundary layers, and the VOF multiphase flow model tracks the free surface of the water-air two-phase flow.

[0053] Specifically, the RANS governing equations use Reynolds-averaged form to describe the continuity and momentum equations of incompressible fluids, and the Reynolds stress terms are closed using the SSTk-ω two-equation turbulence model.

[0054] The RANS governing equations can decompose instantaneous velocity and instantaneous pressure into average and fluctuating quantities; the Reynolds decomposition of instantaneous velocity and instantaneous pressure is expressed as: ; ; in, Instantaneous velocity; Average speed; The pulsation velocity; Instantaneous pressure; Average pressure; The pressure is pulsating. The above formula is used to decompose the instantaneous flow field variables into time-averaged quantities and pulsating quantities, thereby establishing the RANS governing equations.

[0055] Under incompressible fluid conditions, the RANS continuity equation is expressed as: ; in, For the first Average velocity components in each direction; These are the spatial coordinate components. This formula is used to constrain the mass conservation relationship for incompressible fluids.

[0056] Under conditions of an incompressible fluid with approximately constant density, the RANS momentum equation is expressed as: ; in, and The first The and the first Average velocity components in each direction; For time; For fluid density; Average pressure; For the first Spatial coordinate components in each direction; It is the kinematic viscosity coefficient; The average value of the product of pulsating velocities. For the first The pulsating velocity components in each direction For the first The formula describes the pulsating velocity components in each direction. It is used to describe the effects of pressure, viscous diffusion, and Reynolds stress in unsteady average flow.

[0057] The RANS momentum equation described above is one possible expression for the velocity equation. In practice, equivalent stress forms, conservation forms, or discrete forms adapted to the selected finite volume solver can also be used. Equivalent conversions between different expressions can be made based on variable definitions, Reynolds stress term handling methods, and discretization schemes.

[0058] The mean strain tensor is expressed as: ; in, The mean strain tensor; , For the first and the Average velocity components in each direction; , For spatial coordinates. This formula is used to characterize the local deformation rate caused by the average velocity gradient and can be used to illustrate Reynolds stress closure and turbulence model calculations.

[0059] Furthermore, the SSTk-ω two-equation turbulence model uses turbulent kinetic energy... Sum of dissipation To solve for variables, flow separation, backflow, and wall turbulence characteristics are simulated in the locally refined grid and near-wall boundary layer refined grid of the complex water confinement region.

[0060] In one implementation, the turbulent viscosity coefficient in the SSTk-ω two-equation turbulence model is expressed as: ; in, The turbulent viscosity coefficient; For fluid density; These are model constants; It is turbulent kinetic energy; Specific dissipation rate; This represents the absolute value of vorticity. This is the second mixing function in the SSTk-ω two-equation turbulence model; This represents the function that takes the maximum value. This formula is used to illustrate that the turbulent viscosity coefficient can be determined from the turbulent kinetic energy, specific dissipation rate, vorticity, and mixing function.

[0061] The above expression for the turbulent viscosity coefficient is an optional expression of the SSTk-ω two-equation turbulence model. In actual implementation, it can also be set according to the model constants, wall functions, or near-wall treatment methods of the solver used.

[0062] Furthermore, the VOF multiphase flow model is used to track the water-air two-phase flow free surface in the locally refined grid corresponding to the free surface, and the free surface is determined based on the phase volume fraction. Identify the free surface interface mesh.

[0063] The interface determination relationship for phase volume fraction is expressed as follows: ; in, For the first Phase volume fraction of the fluid. When When, it indicates that the corresponding computational grid simultaneously contains the first... For phase fluids and other fluids, the computational grid is defined as a free liquid surface interface grid.

[0064] For a water-air two-phase flow interface mesh, the mixing density is expressed as: ; in, For mixed density; This represents the volume fraction of the first phase fluid. The density of the first phase fluid; The density is the second-phase fluid. The first-phase fluid can be air or water, and the second-phase fluid is the other one; in one embodiment, the first-phase fluid is water and the second-phase fluid is air. This formula illustrates that the density of the interface mesh in the VOF model can be determined by the phase volume fraction.

[0065] Furthermore, the boundary conditions for solving the unsteady flow field include inlet velocity boundary conditions, outlet pressure boundary conditions, wall no-slip boundary conditions, and, when an air domain exists, an air domain outer boundary symmetry condition. The inlet velocity boundary condition constrains the velocity value and direction at the inlet; the outlet pressure boundary condition constrains the pressure at the outlet; the wall no-slip boundary condition constrains at least one wall boundary among ports, wharves, islands, reefs, coastlines, and seabeds; and the air domain outer boundary symmetry condition constrains the velocity gradient at the air domain outer boundary.

[0066] Furthermore, the unsteady flow field solution employs the finite volume method to discretize the flow control equations, and uses a pressure-velocity coupled algorithm to solve the pressure and velocity fields. When the partitioned variable density grid flow field calculation model contains grids with large aspect ratios, grids with large volume differences between adjacent elements, or polyhedral mixed grids, a gradient correction method is used to solve the gradient of the flow field variables to reduce the numerical errors caused by the partitioned variable density grid.

[0067] In practical applications, the partitioned variable density grid flow field calculation model can be solved offline according to the computational resources and mission accuracy requirements. It can also update the local computational domain according to the underwater vehicle's mission area or candidate path area to obtain flow field query results that better meet mission requirements in key navigation areas.

[0068] S7 generates flow field query results.

[0069] Based on the target navigation depth layer of the underwater vehicle, flow field data corresponding to the depth layer, including at least a horizontal velocity component, is extracted from the three-dimensional time-varying velocity field obtained in step S6. Spatial interpolation is then performed on the flow field data to generate flow field query results for underwater vehicle path planning or trajectory tracking control. These flow field query results can be organized according to position coordinates and time indexes, enabling the underwater vehicle to obtain corresponding predicted velocity values ​​at planned path nodes, candidate path nodes, or mission area sampling nodes.

[0070] Specifically, the target navigation depth layer is determined based on the underwater vehicle's mission depth, track tracking depth, or the depth where candidate path nodes are located. The flow field data corresponding to the depth layer includes at least the horizontal velocity component, and may further include vertical velocity components, pressure, or phase volume fraction data as needed for the mission.

[0071] A flow field lookup table can be created based on location coordinates and time indexes. The lookup table can record the predicted flow velocity values ​​at different spatial locations and times within the target navigation depth layer. The query locations can include underwater vehicle planned path nodes, candidate path nodes, or mission area sampling nodes.

[0072] In one embodiment, the data items in the flow field lookup table include horizontal position coordinates. , ,depth ,time X-direction velocity component and the Y-direction velocity component When three-dimensional trajectory tracking or vertical flow field compensation is required, the vertical velocity component can be further included. ,pressure or phase volume fraction .like Figure 5As shown, the target navigation depth layer can be extracted from the three-dimensional time-varying velocity field, and the velocity components within this depth layer can be converted into a flow field lookup table indexed by horizontal position coordinates, target navigation depth, and time.

[0073] When the query location is not located at an existing sampling point or grid node, spatial interpolation can be performed based on the velocity components of the sampling points or grid nodes surrounding the query location; when the query time is not equal to the existing calculation time, temporal interpolation can be performed based on the flow field data of adjacent calculation times. Therefore, the flow field lookup table can provide continuous velocity prediction values ​​for path planning nodes, candidate path nodes, or task area sampling nodes.

[0074] The horizontal velocity component includes an X-direction velocity component and a Y-direction velocity component. Spatial interpolation of the horizontal velocity component can be performed using at least one of cubic spline interpolation, bilinear interpolation, bicubic interpolation, or radial basis function interpolation to generate velocity interpolation prediction results for the corresponding direction. When cubic spline interpolation is used, the piecewise function is expressed as: ; in, For the first A cubic spline function within a segmented interval; , , , For the first spline coefficients within each segmented interval; Let be the spatial location variable. This formula describes one possible spatial interpolation method; in practice, other interpolation methods such as bilinear interpolation, bicubic interpolation, or radial basis function interpolation can also be used.

[0075] In a specific application example, the water area to be modeled is the port entrance and exit area and the nearshore wharf area, which includes the entrance boundary, exit boundary, wharf wall boundary, shoreline boundary, seabed boundary, and free surface. Environmental data includes the port shoreline shape, wharf boundary outline, water depth distribution, inlet flow velocity, inlet flow direction, and initial position of the free surface.

[0076] In this specific application example, the computational domains are divided into complex boundary sub-domains, free surface sub-domains, near-wall boundary layer sub-domains, and main flow sub-domains based on the wharf wall boundary, shoreline curvature, and free surface height range. Specifically, the area near the wharf wall and shoreline is included in the complex boundary sub-domain; the area near the water-air interface is included in the free surface sub-domain; the area near the wharf wall, shoreline, and seabed that meets the near-wall normal distance threshold is included in the near-wall boundary layer sub-domain; and the area far from the aforementioned complex boundaries where flow changes are relatively gentle is included in the main flow sub-domain.

[0077] Locally denser grids with a first grid density are set near the dock, shoreline, near the seabed near the boundary layer, and near the free surface. In the waters far from the wall and located in the main flow area, grids with a second grid density are set for the main flow area. A transitional grid density area is set between the first grid density area and the second grid density area to reduce numerical errors caused by excessive changes in the scale of adjacent grids.

[0078] Interpolation is performed on velocity, pressure, and phase volume fractions at the interface boundaries of adjacent sub-computation domains. Velocity and pressure variables are used for solving the RANS governing equations and pressure-velocity coupling, while phase volume fractions are used for free surface tracking in the VOF multiphase flow model. After completing the unsteady solution, the X-direction and Y-direction velocity components of the underwater vehicle's target navigation depth layer are extracted from the three-dimensional time-varying velocity field. A flow field lookup table indexed by horizontal position coordinates, target navigation depth, and time is generated for use in underwater vehicle path planning or trajectory tracking control.

[0079] like Figure 2 As shown, this embodiment of the invention provides a system for implementing the above method, including an environmental data acquisition module, a three-dimensional model construction module, a sub-computation domain partitioning module, a variable density mesh generation module, an interface variable transfer module, an unsteady flow field solution module, and a flow field result output module.

[0080] The environmental data acquisition module is used to acquire environmental data of the water area to be modeled. The environmental data includes topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and data of complex water area constraints.

[0081] The 3D model building module is used to build a 3D numerical model of the water area to be modeled based on the environmental data. The 3D numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary, and complex water area constraints.

[0082] The sub-computation domain partitioning module is used to divide the computation domain corresponding to the three-dimensional numerical model into multiple sub-computation domains based on the geometric boundaries of the complex water area constraint region, the location of the free liquid surface, the location of the near-wall boundary layer, and the location of the main flow region, and to couple adjacent sub-computation domains through interface boundaries.

[0083] The variable density mesh generation module is used to generate a locally refined mesh with a first mesh density for a sub-computational domain containing geometric boundaries of complex water area constraints, free liquid surfaces, or near-wall boundary layers, and to generate a main flow region mesh with a second mesh density for a sub-computational domain containing the main flow region, wherein the first mesh density is greater than the second mesh density.

[0084] The interface variable transfer module is used to interpolate and transfer flow field variables, including at least velocity and pressure, between adjacent computational domains with different grid densities through interface boundaries, and to interpolate and transfer phase volume fractions when the interface boundary involves a free liquid surface.

[0085] The unsteady flow field solution module is used to solve unsteady flow fields using the RANS governing equations, the SSTk-ω two-equation turbulence model, and the VOF multiphase flow model, to obtain a three-dimensional time-varying velocity field.

[0086] The flow field results output module is used to extract flow field data, including at least the horizontal velocity component, based on the target navigation depth layer of the underwater vehicle, and generate flow field query results for underwater vehicle path planning or trajectory tracking control through spatial interpolation.

[0087] The data flow relationships between the above modules correspond to steps S1 to S7: the environmental data acquisition module outputs environmental data, the 3D model construction module outputs a 3D numerical model, the sub-computation domain partitioning module and the variable density grid generation module output a partitioned variable density grid flow field calculation model, the interface variable transfer module realizes the connection of variables between regions with different grid density, the unsteady flow field solution module outputs a 3D time-varying velocity field, and the flow field result output module outputs flow field query results that can be called by underwater vehicles.

[0088] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the above-described method. The electronic device can be a workstation, server, edge computing device, shipborne or shore-based computing platform, or a computing unit in an underwater vehicle mission planning system. The electronic device can receive aquatic environmental data and perform 3D modeling, mesh generation, unsteady flow field solving, and flow field query result output.

[0089] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described method. The computer-readable storage medium may include a magnetic disk, optical disk, solid-state storage, read-only memory, random access memory, flash memory, or other media capable of storing computer programs.

[0090] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for modeling complex key domain hybrid density grid flow fields for underwater vehicles, characterized in that, include: The environmental data of the water body to be modeled is obtained, including topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and data of complex water body constraint areas; A three-dimensional numerical model of the water area to be modeled is established based on the environmental data, and the three-dimensional numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary and complex water area constraints. Based on the geometric boundaries, free surface location, near-wall boundary layer location, and main flow region location of the complex water area constraint region, the computational domain corresponding to the three-dimensional numerical model is divided into multiple sub-computational domains, and adjacent sub-computational domains are coupled through interface boundaries. A localized mesh with a first mesh density is generated for a sub-computational domain containing the geometric boundary of the complex water area constraint region, the free liquid surface, or the near-wall boundary layer. A main flow region mesh with a second mesh density is generated for a sub-computational domain that is far from the geometric boundary of the complex water area constraint region and located in the main flow region. The first mesh density is greater than the second mesh density. Between adjacent computational domains with different grid densities, flow field variables are interpolated and transferred through interface boundaries, and phase volume fractions are interpolated and transferred when the interface boundaries involve free liquid surfaces, so as to form a partitioned variable density grid flow field computational model. Based on the partitioned variable density grid flow field calculation model, RANS governing equations and SST are used. - The unsteady flow field was solved using the two-equation turbulence model and the VOF multiphase flow model, resulting in the three-dimensional time-varying velocity field of the water body to be modeled. Based on the target navigation depth layer of the underwater vehicle, flow field data corresponding to the depth layer is extracted from the three-dimensional time-varying velocity field, and spatial interpolation is performed on the flow field data to generate flow field query results for underwater vehicle path planning or trajectory tracking control.

2. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The sub-computational domains include a complex boundary sub-computational domain, a free surface sub-computational domain, a near-wall boundary layer sub-computational domain, and a main flow sub-computational domain. The complex boundary sub-computational domain is used to cover local areas near port entrances and exits, wharves, waterways, islands, reefs, coastlines, port boundaries, near-shore structures, or areas with abrupt changes in underwater topography. The free surface sub-computational domain is used to cover areas near the water-air two-phase interface. The near-wall boundary layer sub-computational domain is used to cover areas near the non-slip boundary of the wall. The main flow sub-computational domain is used to cover areas far from complex boundaries and where flow changes are gradual. The division of the sub-computational domains is based on the distance from the geometric boundary of the complex water area constraint region to its spatial location within the computational domain, the curvature of the geometric boundary, the location of the free surface, the normal distance of the near-wall region, and the range of the main flow region.

3. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 2, characterized in that, When the distance from a spatial location within the computational domain to the geometric boundary of a complex water area constraint region is less than a first distance threshold, the spatial location is assigned to a localized denser region corresponding to the first grid density. When the distance is greater than or equal to the first distance threshold and less than the second distance threshold, the spatial location is assigned to a transitional grid density region. When the distance is greater than or equal to the second distance threshold and is located within the main flow region, the spatial location is assigned to the main flow region corresponding to the second grid density. The first distance threshold is less than the second distance threshold, and the first and second distance thresholds are set based on the water area scale, boundary curvature, expected flow field gradient, and computational resources.

4. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The localized mesh includes a localized mesh generated near the free surface, a localized mesh generated at the geometric boundaries corresponding to complex shorelines, islands, reefs, docks, or waterways, and a boundary layer mesh generated near a non-slip wall. The localized mesh is generated using one or more of the following methods: local size constraints, local influence volumes, boundary layer meshes, curvature control meshes, or proximity control meshes. The main flow region mesh uses unstructured meshes, polyhedral meshes, hexahedral meshes, or combinations thereof.

5. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The interpolation transfer of flow field variables through the interface boundary includes: for velocity variables, interpolation mapping is performed based on the positional relationship between the centers of the grid surfaces or grid cells on both sides of the interface, or area-weighted interpolation is performed based on the overlapping area of ​​the interface; for pressure variables, interpolation is performed based on the pressure values ​​of the adjacent cells of the interface, and the interpolation results are used for pressure-velocity coupling solutions; for phase volume fractions, interpolation is performed at the interface boundary involving the free liquid surface, and the interpolated phase volume fractions are subject to range constraints; when the interpolation result is less than 0, it is truncated to 0, and when the interpolation result is greater than 1, it is truncated to 1, or the phase volume fractions in the neighborhood of the interface are normalized so that the volume fractions of each phase satisfy the constraint that they add up to 1; the interface boundary is an interface boundary with matched grid nodes or a non-matched interface boundary with incompletely matched grid nodes.

6. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The SST - Two-equation turbulence model with turbulent kinetic energy Sum of dissipation To solve for variables, the model is used to simulate flow separation, backflow, and wall turbulence characteristics in locally refined grids and near-wall boundary layer refined grids within the complex water-constrained region. The VOF multiphase flow model is used to track the water-air two-phase flow free surface in the locally refined grid corresponding to the free surface, when calculating the phase volume fraction in the grid. satisfy When the calculation grid is defined as a free liquid surface interface grid, the mixing density is determined by weighting the density and volume fraction of each phase fluid for the water-air two-phase flow interface grid.

7. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The boundary conditions for solving the unsteady flow field include inlet velocity boundary conditions, outlet pressure boundary conditions, wall no-slip boundary conditions, and symmetric conditions for the outer boundary of the air domain when an air domain exists. The unsteady flow field is solved by discretizing the flow control equations using the finite volume method and solving the pressure and velocity fields using a pressure-velocity coupled algorithm. When the partitioned variable density grid flow field calculation model contains grids with large aspect ratios, grids with large volume differences between adjacent elements, or polyhedral mixed grids, a gradient correction method is used to solve the gradient of the flow field variables.

8. The method for modeling complex key domain hybrid density grid flow fields for underwater vehicles according to claim 1, characterized in that, The flow field query results are obtained by querying the flow field query table; The flow field query table is built based on location coordinates and time index, and is used to output the predicted flow velocity value at the query location; The query location includes underwater vehicle planned path nodes, candidate path nodes, or mission area sampling nodes; when the query location is not located at an existing sampling point or grid node, spatial interpolation is performed based on the velocity components of the sampling points or grid nodes around the query location; when the query time is not equal to the existing calculation time, temporal interpolation is performed based on the flow field data of adjacent calculation times; the spatial interpolation adopts at least one of cubic spline interpolation, bilinear interpolation, bicubic interpolation, or radial basis function interpolation.

9. A complex water zone variable density grid flow field modeling system for underwater vehicles, characterized in that, The method according to any one of claims 1-8 includes an environmental data acquisition module, a three-dimensional model construction module, a sub-computation domain partitioning module, a variable density mesh generation module, an interface variable transfer module, an unsteady flow field solution module, and a flow field result output module. The environmental data acquisition module is used to acquire environmental data of the water area to be modeled. The environmental data includes topographic boundary data, water depth data, inflow velocity data, inflow direction data, free surface location data, and complex water area constraint region data. The three-dimensional model building module is used to establish a three-dimensional numerical model of the water area to be modeled based on the environmental data. The three-dimensional numerical model includes water topography, inlet boundary, outlet boundary, free surface, wall boundary, and complex water area constraints. The sub-computation domain partitioning module is used to divide the computation domain corresponding to the three-dimensional numerical model into multiple sub-computation domains according to the geometric boundary of the complex water area constraint region, the location of the free liquid surface, the location of the near-wall boundary layer, and the location of the main flow region, and to couple adjacent sub-computation domains through interface boundaries. The variable density mesh generation module is used to generate a locally refined mesh with a first mesh density for a sub-computation domain containing the geometric boundaries of complex water area constraints, free liquid surfaces, or near-wall boundary layers, and to generate a main flow region mesh with a second mesh density for a sub-computation domain containing the main flow region, wherein the first mesh density is greater than the second mesh density. The interface variable transfer module is used to interpolate and transfer flow field variables, including at least flow velocity and pressure, between adjacent sub-computation domains with different grid densities through interface boundaries, and to interpolate and transfer phase volume fraction when the interface boundary involves a free liquid surface. The unsteady flow field solution module is used to solve RANS governing equations and SST. - The unsteady flow field was solved by using the two-equation turbulence model and the VOF multiphase flow model to obtain the three-dimensional time-varying velocity field. The flow field result output module is used to extract flow field data including at least the horizontal velocity component based on the target navigation depth layer of the underwater vehicle, and generate flow field query results for underwater vehicle path planning or trajectory tracking control through spatial interpolation.

10. An electronic device comprising a processor, a memory, and a computer program stored in the memory; characterized in that, When the processor executes a computer program, it implements the complex key domain hybrid density grid flow field modeling method for underwater vehicles as described in any one of claims 1-8.

Images