Outlet vehicle load, motion and multiphase flow field uncertainty quantitative modeling and data analysis method
By using uncertainty quantification technology, the problem of neglecting uncertainty in existing numerical simulations has been solved, enabling accurate quantitative analysis of the load and attitude of the aircraft, and improving the reliability and safety of engineering design.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-11
- Publication Date
- 2026-04-03
AI Technical Summary
Existing numerical simulation analysis methods have failed to effectively quantify uncertainties in shipbuilding and ocean engineering, resulting in insufficient structural design strength reserves and inaccurate attitude prediction, which cannot meet the requirements of modern engineering for accuracy and reliability.
Uncertainty quantification (UQ) technology is used to mathematically represent the uncertainty factors in the simulation input, establish a surrogate model, and perform multi-dimensional uncertainty quantification and analysis, including the calculation of statistical characteristics of single points, curves and cross-sectional positions, and use the Sobol method to analyze the contribution of influencing factors.
It enables the quantification of uncertainties in the load, motion, and multiphase flow field of the vehicle body, provides a more accurate design optimization reference, reduces safety hazards in structural design, and improves the reliability of attitude prediction.
Smart Images
Figure CN121787162A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the fields of shipbuilding and ocean engineering and cross-medium dynamics. Specifically, it relates to a method for quantitative modeling and data analysis of uncertainties in the load, motion, and multiphase flow field of an out-of-water vehicle. Background Technology
[0002] This invention belongs to the interdisciplinary technical fields of computational fluid dynamics, naval architecture and ocean engineering, and cross-medium aircraft dynamics and engineering reliability analysis. Specifically, it relates to a method for quantifying the uncertainty of load, attitude, and flow field results in numerical simulation of underwater vehicles. With the rapid development of computational fluid dynamics and high-performance computing technology, numerical simulation has become a core tool for studying the hydrodynamic characteristics, structural loads, and motion attitude of underwater vehicles. Compared with traditional physical experiments, numerical simulation has many advantages, such as low cost, short cycle time, non-destructive nature, ability to simulate extreme conditions, and provision of detailed information on the entire flow field. It plays an irreplaceable role in the design, optimization, and safety assessment of underwater vehicles.
[0003] However, current mainstream numerical simulation analysis is typically based on deterministic models. That is, during the numerical simulation process, all input parameters (such as incoming flow velocity, angle of attack, turbulence model coefficients, structural stiffness, mass characteristics, etc.) are assumed to be fixed and precise values. Correspondingly, the simulation outputs, such as hydrodynamic loads, vehicle attitude angles, surface pressure distribution, and spatial flow field structure, are also considered as single, deterministic predictions. While this deterministic analysis method is simple and intuitive, its fundamental flaw lies in ignoring the uncertainties that are prevalent in the real world. These uncertainties mainly stem from: (1) Uncertainty of input parameters: Due to measurement errors, manufacturing tolerances, material property dispersion or environmental fluctuations, the true value of the simulation input parameters is not a single point, but exists in a possible range.
[0004] (2) Uncertainty of model form: The physical models on which numerical simulations rely (such as turbulence models, multiphase flow models, material properties, etc.) are themselves simplifications and approximations of complex physical phenomena, and their inherent modeling errors will introduce systematic deviations.
[0005] (3) Uncertainty of numerical discretization: Discretization processes such as grid resolution, time step, and iterative convergence criteria in numerical calculation will also affect the final result.
[0006] In engineering practice, especially in fields with high reliability requirements such as aerospace and shipbuilding and marine engineering, making decisions based solely on deterministic simulation results carries significant risks. For example, a peak load calculated based on deterministic "ideal" input conditions may fail to cover potentially higher loads due to parameter fluctuations, leading to insufficient structural strength reserves and potential safety hazards. Similarly, deterministic predictions of a vehicle's attitude that do not account for uncertainties may mask risks of instability or control failure. Therefore, traditional deterministic analysis is no longer sufficient to meet the stringent requirements of modern engineering for accuracy, reliability, and risk control. To overcome the limitations of deterministic analysis, uncertainty quantification (UQ) techniques have emerged and become a cutting-edge research direction. Summary of the Invention
[0007] The core idea of the UQ method is to mathematically represent the uncertainties in the simulation input (such as probability distribution, number of intervals, etc.) and quantitatively transmit these uncertainties to the simulation output through a systematic mathematical framework, thereby obtaining the statistical characteristics of the uncertainties in the output results (such as load, attitude, flow field information) (such as mean, variance, confidence interval, probability density function, etc.). Based on this, this invention proposes a method for quantitative modeling and data analysis of uncertainties in the load, motion, and multiphase flow field of an outgoing vehicle.
[0008] This invention is achieved through the following technical solution: A method for quantitative modeling and data analysis of uncertainties in the load, motion, and multiphase flow field of an emerging watercraft: The method specifically includes the following steps: S1: For the flow field data of the equidistant spaced vehicle obtained based on the finite volume method, sample data of typical moments are selected and original simulation data of multiple working conditions are collected. S2: Data normalization processing unifies the flow field data coordinates of all cases to eliminate attitude and position deviations, exports standard mesh templates, builds an automated data processing flow, and realizes unified mesh replacement, data reading and batch output of target data for multiple cases; S3: Through data completion, sample space construction, surrogate model establishment and sample generation, statistical characteristic data are calculated, including mean, variance and standard deviation; multi-dimensional uncertainty data quantification and analysis are performed, including for single point location, curve location and cross-sectional location. S4: Based on the statistical characteristic data of step S3, perform curve analysis, mean flow field analysis, variance flow field analysis and sensitivity analysis of the vehicle body, compare the uncertainty of motion attitude of different schemes, calculate the variance contribution of each influencing factor to the target flow field parameters based on the Sobol method and extract the key influencing factors.
[0009] Furthermore, in S1, multiple typical water depth positions based on the characteristic length L of the vehicle are selected, and the preceding case closest to the typical water depth time is found from the simulation iteration file. The calculation time is completed based on the displacement difference. The position deviation is controlled by the completion calculation, and the uncertainty interference caused by the different water depths of the vehicle at the statistical time is reduced.
[0010] Furthermore, in S2, the flow field data of all examples are translated and rotated. The centroid of the vehicle is translated to the geodetic coordinate system (0,0,0), the axis of the vehicle is rotated to a vertical state with a pitch angle of 0°, and the position of the head vertex of the vehicle is translated to (0,0,0). The full three-dimensional flow field data after unifying coordinates is exported through the software interpolation function. The mesh file in this state is exported as a standard mesh template. An automated data processing workflow is constructed, which sequentially executes the following operations: opening the initial flow field file, replacing it with a standard mesh template and scaling the mesh units, reading in the unified 3D flow field interpolation data, saving the updated example file, and outputting the target data. A batch processing mechanism is established, which executes the standardized data processing logic through the automated calling function of the simulation software to complete the batch data processing of multiple examples.
[0011] Furthermore, in S3, the specific process of quantifying and analyzing the uncertainty data of a single point location is as follows: import the single-point data of all cases under the same typical water depth; if a case does not have precise single-point data, find the two nearest time files and obtain the target location data through interpolation; construct a sample space after standardizing all single-point data, and establish a surrogate model based on the sample space; run the surrogate model to obtain a large number of samples, calculate the mean, variance, and standard deviation, and analyze the frequency distribution and probability density distribution.
[0012] Furthermore, in S3, the specific process of quantifying and analyzing the uncertainty data of curve position is as follows: export the specific curve data of all cases under the same typical water depth; for cases with missing curve data, complete the data by interpolation of nearby working conditions and perform standardization processing; establish a surrogate model based on the completed curve samples and generate a large number of samples; calculate the mean and variance of each point of the curve, draw the error band according to the 3σ principle, analyze the propagation characteristics of random factors along the curve direction, and locate the curve area most affected by random factors.
[0013] Furthermore, in S3, the specific process of quantifying and analyzing the cross-sectional position uncertainty data is as follows: The target cross-section is cut in the simulation software, and the flow field data is output in plt format. The plt format data is converted into an ASCII code format dat file, and the phase volume fraction and pressure data of the grid nodes in the dat file are extracted. A surrogate model is established for each grid node in the dat file, and the surrogate model of each node is run to obtain a large number of samples. The mean phase volume fraction, variance of phase volume fraction, and variance of pressure of each node are calculated. The statistical results are substituted into the corresponding nodes in the dat file, and a phase volume fraction mean cloud map and a pressure variance cloud map are generated through visualization software to analyze the high dispersion region of flow field uncertainty.
[0014] Furthermore, in S4, The curve analysis is as follows: Based on the statistical characteristic data of the pitch angle curve of the aircraft output by S3, the mean variation law and 3σ dispersion difference of the pitch angle under the conditions of having and not having the isobaric exhaust scheme are compared, and the optimization effect of the isobaric exhaust technology on the attitude stability of the aircraft during the water-out motion is analyzed. The mean flow field analysis is as follows: the mean phase volume fraction data of each grid node on the symmetry plane output by S3 is substituted into an ASCII code format dat file, and a mean phase volume fraction cloud map is generated by Tecplot visualization software to analyze the average distribution characteristics and thickness law of the air film. The variance flow field analysis is as follows: the volume fraction variance and pressure variance data of each grid node on the symmetry plane output by S3 are substituted into an ASCII code format dat file, and a variance cloud map is generated by Tecplot visualization software to locate regions with high uncertainty dispersion, such as the membrane tail and flow field separation region. The sensitivity analysis is as follows: the flow field pressure near the air film is selected as the target flow field parameter, the platform's 6-DOF motion is taken as the influencing factor, and the contribution ratio of each influencing factor to the total variance of the air film pressure is calculated based on the Sobol method. Among them, the roll motion is the main influencing factor, and the yaw and sway motions are secondary influencing factors.
[0015] A system for quantitative modeling and data analysis of uncertainties in load, motion, and multiphase flow field of an out-of-water vehicle; The system specifically includes an acquisition module, a batch processing module, a multi-dimensional quantization module, and an analysis module: The acquisition module collects original simulation data for multiple operating conditions by selecting sample data at typical moments for the flow field data of the ship at equal intervals obtained based on the finite volume method. The batch processing is used for data normalization, unifying the flow field data coordinates of all cases to eliminate attitude and position deviations, exporting standard mesh templates, constructing an automated data processing flow, and realizing unified mesh replacement, data reading and batch output of target data for multiple cases. The multi-dimensional quantification module calculates statistical characteristic data, including mean, variance, and standard deviation, through data completion, sample space construction, surrogate model establishment, and sample generation. It also performs multi-dimensional uncertainty data quantification and analysis, with the multi-dimensional aspects including single-point location, curve location, and cross-sectional location. The analysis module performs curve analysis, mean flow field analysis, variance flow field analysis, and sensitivity analysis of the vehicle based on the statistical characteristic data of the multi-dimensional quantification module. It compares the uncertainty of motion attitude of different schemes, calculates the contribution of each influencing factor to the target flow field parameters based on the Sobol method, and extracts the key influencing factors.
[0016] An electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the above method.
[0017] A computer-readable storage medium for storing computer instructions that, when executed by a processor, implement the steps of the above-described method.
[0018] Beneficial effects of the invention This invention, based on user-defined random influence conditions, uses non-embedded chaotic polynomial expansion and Monte Carlo methods to establish a reliable surrogate model. It analyzes the propagation characteristics of uncertain random influence factors during the motion of a vehicle, the distribution of phase volume fraction in the flow field, the distribution of pressure in the flow field, and the sensitivity analysis of random influence factors. It can calculate the evolution of target parameters under the influence of random factors, providing a reference direction for design optimization. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of the mesh template exported from the initial calculation example; Figure 2 This is a diagram illustrating how to obtain batch processing jou templates; Figure 3 This is a flowchart for quantifying the uncertainty of data at a specific point location; Figure 4 This is a flowchart for quantifying the uncertainty of data at a specific line position; Figure 5 This is a schematic diagram illustrating the principle of quantifying the uncertainty of the cross-sectional flow field; Figure 6 This is a flowchart for quantifying the uncertainty of the cross-sectional flow field; Figure 7This is an overview of methods for quantifying uncertainties in numerical simulation results of a vehicle emerging from the water. Figure 8 Uncertainty in pitch attitude angle of the emerging watercraft (band diagram); Figure 9 It is the average flow field of phase volume fraction near the exhaust port at position 0.5L; Figure 10 It is the average flow field of phase volume fraction near the exhaust port at position 1.0L; Figure 11 This is a contour plot of the standard deviation of the phase volume fraction near the exhaust port at the 0.5L position; Figure 12 This is a contour plot of the standard deviation of the phase volume fraction near the exhaust port at position 1.0L; Figure 13 It is a pressure variance contour plot near the exhaust port at the 0.5L position; Figure 14 This is a pressure variance contour plot near the exhaust port at the 1.0L position; Figure 15 This is a cloud map of the sensitivity analysis of exhaust gas film pressure. Detailed Implementation
[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0021] Unless otherwise specified, the experimental methods used in the following examples are conventional methods. The materials, reagents, methods, and instruments used, unless otherwise specified, are all conventional materials, reagents, methods, and instruments in the art, and can be obtained commercially by those skilled in the art.
[0022] Combination Figures 1 to 15 This invention proposes a method for performing uncertainty quantification analysis on numerical simulation results of waterborne vehicles, including but not limited to load, motion attitude, and multiphase flow field. For unsteady water discharge processes, in order to maintain relative data comparability, modeling is performed on the influencing parameters of the water discharge process and a specific key result parameter at a certain moment. Considering that the vehicle body is not at the same water depth at the same time under different operating conditions, and the data at different water depths are not comparable due to the flow field environment (relative to uncertainty flow field analysis), the modeling selects data at the same water depth for sample data acquisition, modeling, and uncertainty quantification analysis.
[0023] S1: For the flow field data at equidistant moments obtained based on the finite volume method, sample data at typical moments are selected, and original simulation data for multiple working conditions are collected. For flow field data obtained at equal intervals using the finite volume method, typical time points (e.g., distances of 2L, 1.5L, 1L, 0.5L, 0.0L from the water surface, 0.5L above the water, and 1.0L above the water, where L is the characteristic length of the vehicle) are selected for uncertainty quantification modeling and analysis. The nearest preceding example corresponding to each specific typical time point is found by referencing the files of key mechanical parameters obtained during simulation calculations for each iteration time step. The duration (or iteration time step) of calculation required for the example at the current time point is calculated based on the displacement distance difference between the current time point and the target water depth. The data is calculated to a distance small from the typical target position (e.g., the distance where the control position deviation is less than 0.0001s*Vy), thereby minimizing the uncertainty interference parameters caused by the different water depths of the vehicle at statistical time points.
[0024] S2: Data normalization processing, unifying the flow field data coordinates of all cases to eliminate attitude and position deviations, exporting standard mesh templates, and compiling automated data processing scripts; Considering that samples under different variables may not be precisely controlled to the same displacement for the flow field processing data at the same time, or due to iterative calculation deviations, or because the data from different examples obtained using mesh reconstruction calculations for a certain local mesh point, a certain line mesh point, or a certain cross-section mesh point cannot be compared, read, or processed due to the inability to compare, read, and process the derived mesh count, mesh coordinates, and mesh coordinate file sorting, it is impossible to model specific points based on the data with quantified uncertainty. The following method is adopted for data normalization processing.
[0025] For different calculation cases (e.g., using full-rank collocation for three variables to study the effects of different velocities, air chamber pressures, and water depths; in this case, there are 20 calculation cases, each with a different number of time-time files (e.g., flow field data at 100 time-times)), after processing the data at the same typical water depth time-time in step 1, the model flow field data is translated and rotated for each calculation case: the vehicle's center of mass is translated to the geodetic coordinate system (0,0,0) → the vehicle's axis is rotated to a vertical position (pitch angle = 0°) → the vehicle's nose vertex is translated to (0,0,0). All three-dimensional flow field data is exported using the software interpolation function. This data is named: **L.ip Export the initial mesh file of the numerical simulation example (i.e., the mesh file when the model's head vertex is located at (0,0,0) and the vehicle's attitude angle is 0°) and record the file name as **-start.mesh. The process is shown in Figure 1.
[0026] The following functions are recorded and executed using the jou script file: Open the first flow field file of this calculation case (i.e., the calculation case file when the model head vertex is located at (0,0,0) and the vehicle attitude angle is 0°) → Replace the mesh with: **-start.mesh → scal mesh unit is mm → Read the interpolation file **L.ip → Save the calculation case file → Output the required flow field data of the curve at a specific position of the vehicle (such as the pressure at the position of the vehicle's generatrix) → Output the required flow field data of the specific cross-section position of the vehicle (such as the pressure, phase, velocity, etc. at the position of the symmetry plane) → Output other required data, and record the jou file name as **-output.jou. The process is shown in Figure 2.
[0027] Use the batch processing file (.bat) built into Windows to create a batch processing file to achieve the following function: Open Fluent → d to read **-output.jou → Exit Fluent; S3: Multi-dimensional uncertainty quantification modeling and statistical characteristic acquisition, respectively targeting single point location, curve location and cross-sectional location, through data completion, sample space construction, surrogate model establishment and sample generation, to calculate statistical characteristic data, including mean, variance, standard deviation and basic data required for visualization; For the quantification and analysis of uncertain data at a single point location (e.g., the pressure at the center of the bottom of a vehicle at a distance of 1.0L above the water - which needs to be exported in ASCII format): Import data from different examples at a specific typical water depth location (e.g., the pressure data at the center of the bottom of the vehicle) into MATLAB → If there is no working condition at this precise spatial location, find the two nearest process files and perform interpolation → After dimensionless transformation of all data, establish a sample space → Build a surrogate model based on the sample space and run it 1 million times to obtain a large number of samples → Obtain statistical characteristics such as mean, variance, and standard deviation, and analyze frequency distribution, probability density distribution, etc. The process is shown in Figure 3.
[0028] For quantification and analysis of uncertain data regarding the position of a curve (e.g., the uncertainty distribution of axial load on a vehicle under a certain operating condition): Output data on a specific curve → Import data from different examples at a specific typical water depth position using MATLAB → If there is no operating condition at that precise spatial position, interpolation needs to be performed on the two nearest process files → After dimensionless transformation of all data, establish a sample space → Build a surrogate model based on the sample space and run it 1 million times to obtain a large number of samples → Calculate the standard deviation and then, based on 3... σ The principle is to draw error bands → analyze the propagation characteristics of random influencing factors along the axial direction based on the trend of the error bands, and optimize the positions that are more affected by random factors. The process is shown in Figure 4.
[0029] For quantification and analysis of uncertain cross-sectional location data (such as symmetric flow field data of a vehicle at a distance of 1.0L above the water): Cut out the cross-section to be processed in Fluent and output it as a plt format file → Open the plt format file with Tecplot and output an ASCII code format dat file → In the ASCII edit file, the phase volume fraction data and pressure data on the grid nodes can be located according to the data format → Extract the data from all sample grid nodes and build a surrogate model on each node, as shown in Figure 5 → Run the surrogate model 1 million times to obtain a large number of samples → Calculate the mean, variance, and pressure variance of the phase volume fraction on each grid → Analyze the areas with greater dispersion when the flow field is affected by random conditions based on the cloud map, as shown in Figure 6.
[0030] The overall uncertainty analysis and post-processing flow of the numerical simulation results of the water-emerging motion of the vehicle is shown in Figure 7.
[0031] S4: Uncertainty Impact Patterns and Sensitivity Analysis. Based on the statistical characteristic data from step S3, conduct curve analysis examples, mean flow field analysis, and variance flow field analysis. Compare the uncertainty patterns of motion attitudes of different schemes. Calculate the variance contribution of each influencing factor to the target flow field parameters using the Sobol method and extract the key influencing factors.
[0032] Example of curve analysis: Quantitative analysis of the uncertainty of pitch attitude during the water exit motion of a vehicle. As shown in Figure 8, the blue curve represents the pitch angle variation and dispersion of a conventional vehicle emerging from the water, while the red curve represents the pitch angle variation and dispersion of a vehicle equipped with isobaric exhaust stabilization technology. The curves represent the mean pitch angle. From this perspective, the initial attitude changes of vehicles with and without exhaust are almost identical. However, the attitude change of vehicles with exhaust gradually converges to stability, while the attitude of vehicles without exhaust continues to increase and deteriorate, showing no convergence trend. The width of the color bands is determined according to 3... σ The diagram, plotted according to principle, represents the magnitude of the data dispersion at that point. Analysis of the graph shows that the pitch attitude of a vehicle with exhaust gas initially increases and then converges under random conditions, while the pitch attitude dispersion of a vehicle without exhaust gas continuously increases without convergence. The information in the graph indicates that isobaric exhaust gas effectively reduces the pitch attitude of a vehicle, resulting in a stabilizing effect. This effect requires a certain time delay; under the influence of the air film, the pitch deflection of the vehicle converges to a stable level, and the dispersion is significantly reduced compared to a vehicle without exhaust gas.
[0033] Mean flow field analysis; mean distribution of exhaust film phase volume fraction of the vehicle body As shown in Figures 9 and 10, 0.5L The flow field plot with mean phase volume fraction at location and 1.0 L Comparing the mean flow field diagram of phase volume fraction at the location, the mean gas film size expands spatially and gradually covers most of the circumferential surface of the model. The circumferentially unfused strip-shaped gas films form an approximately isobaric region of a certain size. From the mean streamline distribution inside the gas film, similar to the results obtained under deterministic emission conditions, a saddle-point flow structure appears at the separation point between the gas phase region and the mixed phase region. The unfused gas film at the tail of the film exhibits a closed form of saddle-point-double helix junction.
[0034] Variance flow field analysis: Variance distribution characteristics of exhaust film pressure and phase volume fraction of the outgoing water-carrying body As shown in Figures 11 and 13, for motion 0.5 L Distribution of standard deviations, standard deviation of phase volume fraction σ α The main concentrations are in the mixed phase region (②) and at the gas-water interface at the edge of the gas film. From within the symmetry plane... σ α The distribution shows that the standard deviation gradually increases as the axial position extends downstream, indicating the uncertainty in the size of the exhaust film space under random launch conditions. The pressure standard deviation is shown in the figure. σ p The distribution shows that high σ p The concentration is near the exhaust port, indicating a high degree of uncertainty in exhaust intensity under random emission conditions. Furthermore, because the exhaust film tail is relatively thin at this point, a strong retroflow has not yet formed, resulting in a high standard deviation of the film tail pressure. σ p The values were not high.
[0035] As shown in Figures 12 and 14, for motion 1.0 L The distribution of standard deviations is determined by the standard deviation of phase volume fractions. σ α The distribution shows that high σ α The numerical values continue to converge on both sides of the bubble boundary in the latter half of the mixed phase region, representing the uncertainty of the bubble's spatial propulsion distance. Furthermore, the position of the central axis of the reverse flow trajectory of the reflected flow can also be observed. σ α The values are low, with multiple high-value areas appearing only on the surface of the vehicle body at the termination position and within the radial symmetry plane. σ α This region reflects the uncertainty in the reverse flow distance and water-carrying capacity of the water-absorbing phase as it enters the membrane. For the pressure standard deviation... σ pThe distribution of the initial-scale air film membrane tail closes on the surface of the vehicle body, generating local stagnation high pressure. At this time, the uncertainty of the membrane tail closure strength and position caused by random launch conditions makes the pressure uncertainty accumulate near the membrane tail, which manifests as multiple isolated high pressure standard deviation regions.
[0036] Sensitivity analysis As shown in Figure 15, the sensitivity of the flow field near the isobaric exhaust film at position 2.0L to the platform's 6-DOF motion was analyzed using the Sobol sensitivity analysis method. The results are shown in Figure 15. The total variance of the pressure distribution within the film mainly originates from the roll, yaw, and sway motions, with the roll motion contributing the most to the variance. The platform's motion velocity, vertical velocity, and pitch motion have minimal impact on the variance of the pressure within the film, with the vertical velocity and pitch motion contributing almost nothing.
[0037] A system for quantitative modeling and data analysis of uncertainties in load, motion, and multiphase flow field of an out-of-water vehicle; The system specifically includes an acquisition module, a batch processing module, a multi-dimensional quantization module, and an analysis module: The acquisition module collects original simulation data for multiple operating conditions by selecting sample data at typical moments for the flow field data of the ship at equal intervals obtained based on the finite volume method. The batch processing is used for data normalization, unifying the flow field data coordinates of all cases to eliminate attitude and position deviations, exporting standard mesh templates, constructing an automated data processing flow, and realizing unified mesh replacement, data reading and batch output of target data for multiple cases. The multi-dimensional quantification module calculates statistical characteristic data, including mean, variance, and standard deviation, through data completion, sample space construction, surrogate model establishment, and sample generation. It also performs multi-dimensional uncertainty data quantification and analysis, with the multi-dimensional aspects including single-point location, curve location, and cross-sectional location. The analysis module performs curve analysis, mean flow field analysis, variance flow field analysis, and sensitivity analysis of the vehicle based on the statistical characteristic data of the multi-dimensional quantification module. It compares the uncertainty of motion attitude of different schemes, calculates the contribution of each influencing factor to the target flow field parameters based on the Sobol method, and extracts the key influencing factors.
[0038] An electronic device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the above method.
[0039] A computer-readable storage medium for storing computer instructions that, when executed by a processor, implement the steps of the above-described method.
[0040] The memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM). It should be noted that the memory of the methods described in this invention is intended to include, but is not limited to, these and any other suitable types of memory.
[0041] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means such as coaxial cable, optical fiber, digital subscriber line, DSL, or wireless means such as infrared, wireless, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium such as a floppy disk, hard disk, magnetic tape; an optical medium such as a high-density digital video disc, DVD; or a semiconductor medium such as a solid-state disk, SSD, etc.
[0042] In implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software. The steps of the method disclosed in the embodiments of this application can be directly implemented by a hardware processor, or by a combination of hardware and software modules in the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, detailed descriptions are omitted here.
[0043] It should be noted that the processor in the embodiments of this application can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiments can be completed by the integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied as execution by a hardware decoding processor, or as execution by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above methods.
[0044] The foregoing has provided a detailed description of the method for quantifying and analyzing the uncertainties of load, motion, and multiphase flow field of an outgoing vehicle proposed in this invention. The principles and implementation methods of this invention have been explained. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.
Claims
1. A method for quantitative modeling and data analysis of uncertainties in the load, motion, and multiphase flow field of an out-of-water vehicle, characterized in that: The method specifically includes the following steps: S1: For the flow field data of the equidistant spaced vehicle obtained based on the finite volume method, sample data of typical moments are selected and original simulation data of multiple working conditions are collected. S2: Data normalization processing unifies the flow field data coordinates of all cases to eliminate attitude and position deviations, exports standard mesh templates, builds an automated data processing flow, and realizes unified mesh replacement, data reading and batch output of target data for multiple cases; S3: Through data completion, sample space construction, surrogate model establishment and sample generation, statistical characteristic data are calculated, including mean, variance and standard deviation; multi-dimensional uncertainty data quantification and analysis are performed, including for single point location, curve location and cross-sectional location. S4: Based on the statistical characteristic data from step S3, perform curve analysis, mean flow field analysis, variance flow field analysis, and sensitivity analysis of the vehicle body. Compare the uncertainty of motion attitude under different schemes. Calculate the variance contribution of each influencing factor to the target flow field parameters based on the Sobol method and derive the key influencing factors.
2. The method according to claim 1, characterized in that: In S1, multiple typical water depth positions are selected based on the characteristic length L of the vehicle body. The preceding case closest to the typical water depth time is found from the simulation iteration file. The calculation time is completed based on the displacement difference. The position deviation is controlled by the completion calculation, and the uncertainty interference caused by the different water depths of the vehicle body at the statistical time is reduced.
3. The method according to claim 2, characterized in that: In S2, the flow field data of all examples are translated and rotated. The centroid of the vehicle is translated to the geodetic coordinate system (0,0,0), the axis of the vehicle is rotated to a vertical state with a pitch angle of 0°, and the position of the head vertex of the vehicle is translated to (0,0,0). The full three-dimensional flow field data after unifying coordinates is exported through the software interpolation function. The mesh file in this state is exported as a standard mesh template. An automated data processing workflow is constructed, which sequentially executes the following operations: opening the initial flow field file, replacing it with a standard mesh template and scaling the mesh units, reading in the unified 3D flow field interpolation data, saving the updated example file, and outputting the target data. A batch processing mechanism is established, which executes the standardized data processing logic through the automated calling function of the simulation software to complete the batch data processing of multiple examples.
4. The method according to claim 3, characterized in that: In S3, the specific process of quantifying and analyzing single-point location uncertainty data is as follows: import single-point data of all cases under the same typical water depth; if a case does not have precise single-point data, find the two nearest time files and obtain the target location data through interpolation; construct a sample space after standardizing all single-point data, and establish a surrogate model based on the sample space; run the surrogate model to obtain a large number of samples, calculate the mean, variance, and standard deviation, and analyze the frequency distribution and probability density distribution.
5. The method according to claim 4, characterized in that: In S3, the specific process of quantifying and analyzing the uncertainty of curve position data is as follows: export the specific curve data of all cases under the same typical water depth; for cases with missing curve data, complete the data by interpolation of nearby working conditions and perform standardization processing; establish a surrogate model based on the completed curve samples and generate a large number of samples; calculate the mean and variance of each point of the curve, draw the error band according to the 3σ principle, analyze the propagation characteristics of random factors along the curve direction, and locate the curve area most affected by random factors.
6. The method according to claim 5, characterized in that: In S3, the specific process of quantifying and analyzing the cross-sectional position uncertainty data is as follows: cut the target cross-section in the simulation software and output the flow field data in plt format, convert the plt format data into ASCII code format dat file, and extract the phase volume fraction and pressure data of the grid nodes in the dat file. A proxy model is established for each grid node in the dat file. The proxy model of each node is run to obtain a large number of samples. The mean of phase volume fraction, variance of phase volume fraction, and variance of pressure are calculated for each node. The statistical results are substituted into the corresponding nodes in the dat file, and the mean of phase volume fraction and variance of pressure are generated by visualization software to analyze the region with high dispersion of flow field uncertainty.
7. The method according to claim 6, characterized in that: In S4, The curve analysis is as follows: Based on the statistical characteristic data of the pitch angle curve of the aircraft output by S3, the mean variation law and 3σ dispersion difference of the pitch angle under the conditions of having and not having the isobaric exhaust scheme are compared, and the optimization effect of the isobaric exhaust technology on the attitude stability of the aircraft during the water-out motion is analyzed. The mean flow field analysis is as follows: the mean phase volume fraction data of each grid node on the symmetry plane output by S3 is substituted into an ASCII code format dat file, and a mean phase volume fraction cloud map is generated by Tecplot visualization software to analyze the average distribution characteristics and thickness law of the air film. The variance flow field analysis is as follows: the volume fraction variance and pressure variance data of each grid node on the symmetry plane output by S3 are substituted into an ASCII code format dat file, and a variance cloud map is generated by Tecplot visualization software to locate regions with high uncertainty dispersion, such as the membrane tail and flow field separation region. The sensitivity analysis is as follows: the flow field pressure near the air film is selected as the target flow field parameter, the platform's 6-DOF motion is taken as the influencing factor, and the contribution ratio of each influencing factor to the total variance of the air film pressure is calculated based on the Sobol method. Among them, the roll motion is the main influencing factor, and the yaw and sway motions are secondary influencing factors.
8. A system for quantitative modeling and data analysis of uncertainties in the load, motion, and multiphase flow field of an out-of-water vehicle, characterized in that: The system is used to perform the steps of the method according to any one of claims 1 to 7; The system specifically includes an acquisition module, a batch processing module, a multi-dimensional quantization module, and an analysis module: The acquisition module collects original simulation data for multiple operating conditions by selecting sample data at typical moments for the flow field data of the ship at equal intervals obtained based on the finite volume method. The batch processing is used for data normalization, unifying the flow field data coordinates of all cases to eliminate attitude and position deviations, exporting standard mesh templates, constructing an automated data processing flow, and realizing unified mesh replacement, data reading and batch output of target data for multiple cases. The multi-dimensional quantification module calculates statistical characteristic data, including mean, variance, and standard deviation, through data completion, sample space construction, surrogate model establishment, and sample generation. It also performs multi-dimensional uncertainty data quantification and analysis, with the multi-dimensional aspects including single-point location, curve location, and cross-sectional location. The analysis module performs curve analysis, mean flow field analysis, variance flow field analysis, and sensitivity analysis of the vehicle based on the statistical characteristic data of the multi-dimensional quantification module. It compares the uncertainty of motion attitude of different schemes, calculates the contribution of each influencing factor to the target flow field parameters based on the Sobol method, and derives the key influencing factors.
9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method of claim 8.
10. A computer-readable storage medium for storing computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the method of claim 8.