Rapid design method for pedigree appearance of underwater vehicle

Abstract

The invention discloses a rapid design method for a pedigree profile of an underwater vehicle. The method comprises the following steps: S1, establishing a parameterized profile of the underwater vehicle; s2, establishing a hydrodynamic numerical simulation analysis method of the appearance, wherein the hydrodynamic numerical simulation analysis method comprises appearance grid division and solver setting; s3, establishing a motion characteristic evaluation index of the underwater vehicle; s4, sampling and analyzing to obtain appearance parameters and hydrodynamic response thereof; s5, establishing a high-precision hydrodynamic agent model by using the data set, and analyzing a sensitivity relationship between the appearance parameters and the motion characteristics; and S6, solving an optimization problem with constraint conditions through an optimization algorithm to obtain optimal geometric parameters of the appearance. The method provided by the invention breaks through the limitation of the traditional empirical method, ensures that the method can quickly determine the geometric parameters of the shape of the underwater vehicle according to different underwater observation and detection task requirements, shortens the design period of the shape of the underwater vehicle, and promotes the lineage design of the underwater vehicle.

Description

Technical Field

[0001] The present invention relates to the field of underwater vehicle shape optimization design, and in particular to a method for rapid design of underwater vehicle shape based on genealogy Background Art The ocean has a profound impact on human production and life. In order to better understand and utilize the ocean, countries have invested a lot of resources to build various types of ocean observation equipment, including satellites and underwater vehicles. The application of these devices not only helps us understand the ocean environment, climate change and ecosystem, but also promotes the development of fisheries, shipping, national defense and other fields. With the increasing complexity and scale of ocean observation tasks, the role of underwater vehicles is becoming increasingly important and has become an indispensable tool in ocean exploration. Therefore, researching and improving the design and performance of underwater vehicles has become an important direction of current marine science and technology.

[0002] In the design of underwater vehicles, the shape unit is the core factor affecting its multi-dimensional motion characteristics. Underwater vehicles need to perform various tasks in complex marine environments. Their shape not only directly affects the endurance, speed, stability and other motion performance during navigation, but also has a coupling effect with structural design, energy system, mission layout and other aspects. Therefore, shape design is not an isolated process, but a comprehensive design task that is interrelated and mutually influenced by multiple design units. In order to meet different marine observation needs, the shape design of underwater vehicles usually needs to balance different performance indicators such as endurance, load mass, and diving depth. For example, when designing the shape of a vehicle, it may be necessary to first determine its diving depth, and then optimize its shape to improve the navigation economy under the premise of meeting specific mass and maneuverability requirements. This optimized design enables underwater vehicles to perform tasks more efficiently in complex and changing marine environments.

[0003] The traditional underwater vehicle design method mainly relies on the experience of designers, determines the initial parameters through the design models in previous projects, and then combines computational fluid dynamics (CFD) and other means to perform numerical verification and gradual adjustment. However, this method has obvious disadvantages. First, excessive reliance on experience limits innovation, and designers often find it difficult to break through previous design ideas, resulting in the inability to explore better solutions. Secondly, the empirical method can usually only provide a preliminary design direction. Designers need to repeatedly optimize through a large number of numerical simulations and experiments, resulting in extremely large consumption of computing resources and time, and increasing design costs. In addition, due to the lack of systematic optimization methods in traditional methods, it is difficult to find the truly optimal design under multi-dimensional performance requirements, which makes the final solution may only be a "suboptimal solution" that meets the needs, rather than a truly optimal solution, and it is difficult to give full play to the potential of underwater vehicles. Summary of the invention

[0004] In response to the problems existing in the prior art, the present invention provides a method for rapid design of the systematic shape of an underwater vehicle, which solves the problems that the existing method for designing the shape of an underwater vehicle is too dependent on experience, has low design accuracy and efficiency, is difficult to adapt to the growing requirements of ocean observation and detection tasks, and cannot obtain the optimal parameters of the shape.

[0005] The present invention is implemented as follows: a rapid design method for the pedigree shape of an underwater vehicle comprises the following steps: S1. Establish a parameterized model of underwater vehicle; S2. Establish a hydrodynamic numerical simulation analysis model for the shape, including shape mesh division and solver settings; S3. Establish evaluation indexes for underwater vehicle motion characteristics; S4. Sampling and analysis to obtain shape parameters and their hydrodynamic response; S5. Use the data set to build a high-precision proxy model and analyze the sensitivity relationship between shape parameters and motion characteristics; S6. Solve the optimization problem with constraints through optimization algorithm to obtain the optimal geometric parameters of the shape.

[0006] In the present technical solution, preferably, the step S1 comprises: determining the basic form of the shape according to the design requirements of the underwater vehicle shape, and establishing an initial three-dimensional shape model using Catia software.

[0007] In the present technical solution, preferably, the step S2 includes the following steps: importing the three-dimensional model into the ICEM software, setting the angle of attack, sideslip angle of the underwater vehicle and the fluid domain in which it is located, and then performing grid division; importing the obtained grid file into the Fluent software, selecting the pressure-based solver, selecting the k-ω SST turbulence model, setting the boundary type and incoming flow velocity boundary conditions, and selecting the pressure discrete format as Coupled; and iteratively solving and calculating after initialization.

[0008] In the present technical solution, preferably, step S3 includes: establishing an underwater vehicle dynamics model based on the Newton-Euler method, and analyzing the stability, navigation economy, navigation speed and turning maneuverability of the underwater vehicle dynamics model.

[0009] In the present technical solution, preferably, the step S4 includes the following steps: integrating the steps S1 and S2, determining the shape parameters to be optimized in step S1, the angle of attack and the sideslip angle in step S2 as input variables, and taking the resistance of the underwater vehicle as input variables. F D , lift F L , lateral force F SF , rolling moment Mrol , yaw moment M yaw and pitch moment M pit is the output variable; the sampling test is designed using the optimized Latin hypercube method, and the output response of the sample is calculated using the hydrodynamic numerical analysis method.

[0010] In the present technical solution, preferably, step S5 includes the following steps: using the above-mentioned data set, using the cross-validation method, comparing the accuracy of different proxy model training methods to establish proxy models of various hydrodynamics, and selecting the model training method with the highest accuracy. Then the data set is evenly divided into 5 groups, 4 of which are training sets, and the remaining group of data is used as a validation set. Five groups of proxy models are trained respectively, and the data with poor prediction results in the five groups of proxy models are deleted, and the proxy models are retrained. Using the motion characteristic evaluation index established in step S2, the sensitivity of each design variable is analyzed using the Sobol' sensitivity analysis method, which is used to optimize the selection of variables before optimization. Variables with lower sensitivity can be adjusted according to the needs of other design units.

[0011] In the present technical solution, preferably, step S6 includes the following steps: according to actual design requirements, comprehensively considering the volume, stability, navigation economy, navigation speed and turning maneuverability of the underwater vehicle, setting constraints and optimization goals, and finding the optimal size design scheme of the appearance under specific working conditions.

[0012] The present invention proposes a method for rapid design of underwater vehicle pedigree appearance, which realizes rapid optimization of underwater vehicle appearance design through systematic and data-driven means, and has multiple significant advantages and beneficial effects.

[0013] First, the method establishes the relationship between the underwater vehicle's multidimensional motion characteristics and its hydrodynamic coefficients. The stability, turning maneuverability, navigation economy and speed of the underwater vehicle can be analyzed according to the set control parameters. This multidimensional analysis can more comprehensively reflect the performance of the underwater vehicle in actual navigation and ensure that the design can cope with complex marine environments. Through this relationship model, designers can more accurately predict the performance of the vehicle under different conditions, thereby optimizing its various motion characteristics and improving its overall performance in the early stages of design.

[0014] Secondly, this method establishes a parametric model of the underwater vehicle, which can simultaneously consider the design parameters of the fuselage and the appendages. This parametric model not only covers the overall structure of the vehicle, but also considers the mutual influence between the various parts, which makes it possible to obtain the global optimal solution during the design optimization process, rather than simply optimizing a certain part. This global optimization method ensures that the underwater vehicle can maximize its overall efficiency and functionality while meeting various performance requirements.

[0015] Third, based on the data-driven design method, this method uses a back-propagation neural network model to determine the relationship between the underwater vehicle's shape parameters and hydrodynamic performance (such as drag, lift, lateral force, rolling moment, yaw moment, and pitch moment). The application of this neural network proxy model effectively overcomes the problem of the traditional method that it is difficult to accurately describe the fluid motion characteristics with empirical formulas, making the underwater vehicle's shape design more scientific and accurate. At the same time, through the training and verification of a large number of data sets, the model can greatly reduce the reliance on fluid mechanics calculations, thereby speeding up the design process.

[0016] Fourth, the method proposed in the present invention has wide applicability, especially when the shape parameters vary greatly, it can still stably optimize the design. The data-driven model established by this method is applicable to a variety of working conditions, and can not only handle conventional design requirements, but also cope with more complex design scenarios. This widely applicable feature enables the design method to quickly give optimized shape parameters under different mission requirements and operating environments without repeating complex hydrodynamic calculations, thereby greatly shortening the design cycle and improving efficiency.

[0017] Finally, the design method is also highly versatile. Whether it is the shape design of other types of underwater vehicles or the shape optimization of underwater vehicles using different parameterization methods, this method can provide good applicability and performance improvement. This versatility makes the method not only applicable to current design tasks, but also has the potential to be expanded to meet diverse design needs in the future, and has broad application prospects.

[0018] In general, the underwater vehicle genealogical shape rapid design method of the present invention can efficiently and accurately complete the underwater vehicle shape design through data-driven and parametric optimization means, which significantly improves the design efficiency and performance optimization level. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the overall process of the method of the present invention; Figure 2 It is a schematic diagram of a parameterized model of an underwater vehicle in an application example of the method of the present invention; Figure 3 A schematic diagram of the fluid domain of an underwater vehicle in an application example of the method of the present invention; Figure 4-7 It is a schematic diagram of the change of geometric parameters of the underwater vehicle before and after optimization in the application example of the method of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0021] In order to solve the problem that the existing underwater vehicle shape design method is too dependent on experience, has low design accuracy and efficiency, is difficult to adapt to the growing requirements of ocean observation and detection tasks, and cannot obtain the optimal shape parameters, the present invention provides a rapid design method for the pedigree shape of an underwater vehicle. In order to further illustrate the structure of the present invention, the detailed description is as follows in conjunction with the accompanying drawings: like Figure 1 As shown in the figure, the rapid design of underwater vehicle pedigree shape is used to analyze the shape of a certain type of underwater vehicle, including: S1. Determine the basic form of the underwater vehicle, that is, establish a parameterized model of the underwater vehicle: The shape of an underwater vehicle largely determines its motion characteristics and the design of its pressure hull. In order to ensure that the underwater vehicle performs well in terms of stability and maneuverability, and to maximize the endurance economy, it is crucial to optimize the shape design.

[0022] In this embodiment, select Figure 2 The parametric model of the underwater vehicle shown in the figure is used as an example to quickly design the pedigree shape, in which the main body line type of the fuselage is the Myring line type, and its expression is:

[0023] in, x It is the axial distance from a certain point on the shape to the tip of the nose. a , b , c Respectively represent the length of the bow, middle and tail of the underwater glider, n and θ t They represent the nose sharpness factor and tail departure angle of the underwater glider respectively.

[0024] The ranges defining the fuselage design parameters are: diameter d ∈[100, 500], unit is mm; header length a = f ba · d , where the head aspect ratio f ba ∈[1, 3]; middle length b = f bb · d , where the middle aspect ratio fbb ∈[3, 8]; tail length c = f bc · d , where the tail aspect ratio f bc ∈[1, 3]; Head sharpness factor n ∈[0.5, 3]; tail departure angle θ t ∈[0, 30], unit: °. The range of wing design parameters is defined as: wing half span ws=L B · f ws , where the wing half-aspect ratio f ws ∈[0.2, 0.4]; wing root length wr=f wr · b , where the wing root diameter ratio f wr ∈[0.1, 0.4]; wing position according to the nose wp= f wp ·( b-wr ), where the wing position coefficient f wp ∈[0.5, 0.9]; wing tip length wt=f wt · wr , where the wing tip root ratio f wt ∈[0.2, 1]; wing leading edge angle wsa ∈[30, 90], unit: °. The range of tail design parameters is defined as: tail half span ss=f ss · d , where the tail half aspect ratio f ss ∈[0.6, 1.5]; tail wing root length sr=f sr · c , where the tail root diameter ratio f sr ∈[0.3, 0.6]; tail wing is located at the head position sp=f sp ·( c-sr ), where the tail position coefficient f sp ∈[0.2,0.8]; tail wing tip length st=f st · sr , where the tail tip root ratio fwt ∈[0.2, 1]; tail wing trailing edge angle ssa ∈[30,90], unit: °.

[0025] In this embodiment, the Myring line shape is selected as the fuselage main body line shape, and multiple design parameters are defined. The main parameters of the fuselage main body include the diameter, the length of the bow, the middle part, and the tail, as well as the bow sharpness factor and the tail departure angle. The definition of these parameter ranges lays the foundation for the subsequent optimization of the underwater vehicle shape.

[0026] S2. Establish a hydrodynamic numerical simulation analysis model for the shape, including shape mesh division and solver settings.

[0027] The underwater vehicle shape model and the outer boundary of its fluid domain are determined by using Catia software. The steps include: linking the geometric shape parameters that need to be changed in the .CATPart file to the design table .txt file, closing and reopening the Catia software, and using a macro file to record the following operations: (1) changing the shape geometric parameters in the .txt file; (2) opening the original .CATPart file; (3) updating the geometry file; (4) saving the geometry file; (5) closing the window. The recorded macro file is used for subsequent batch data collection. In order to determine the position derivative and rotation derivative of the underwater vehicle coefficient, it is necessary to perform numerical simulation of the towing tank and the spiral arm tank. The outer boundary of the fluid domain is as follows: Figure 3 Its related parameters are shown in Table 1.

[0028] Table 1. Fluid domain related parameters

[0029] Similarly, perform unstructured meshing on the fluid domain and record a macro file. Use the File-Import-model command to import the .CATPart file into the ICEM software. Then define the PART names such as IN, OUT, WALL and their corresponding geometric elements so that boundary conditions can be set in the turbulence solver Fluent later. Rotate the underwater vehicle to a certain angle of attack and sideslip, where the angle of attack is α ∈[0, 10], unit is °, sideslip angle β ∈[0, 10], unit is °. Then the model is geometrically topologically constructed, and the geometric body between the outer boundary and the underwater vehicle is generated, namely the fluid domain WATER. The volume mesh and surface mesh sizes in the fluid domain are set to 512 mm and 8 mm, and the surface mesh sizes of some large areas are set to 16 mm. A prismatic layer mesh is set for the surface of the underwater vehicle, with a first layer height of 0.2 mm, a total number of layers of 15, and a mesh growth rate of 1.15. After the parameters are set, an unstructured mesh is generated, and the mesh is exported to .msh format after smoothing.

[0030] Use Fluent TUI script command to import the mesh and scale it, then define its turbulence model with curvature-corrected k-ω SST model. Set the medium of the fluid domain to water and define its density to 1021 kg / m 3 . Set the boundary conditions of the fluid domain in sequence. In the numerical simulation of the towing pool, the boundary condition of the fluid inlet IN is the velocity inlet, and the velocity magnitude is 0.5 m / s; in the numerical simulation of the spiral arm pool, the boundary condition of the fluid inlet IN is the velocity inlet, and the velocity magnitude is 0.5 m / s. R d = 20 m Set the rotation center position coordinates in the cylindrical coordinate system, and the angular velocity is 0.025 rad / s; the boundary condition of the outlet OUT is free outflow; the boundary condition of the outer wall WALL of the fluid domain is symmetric boundary condition; the surface of the underwater vehicle is the wall, such as Figure 3 As shown. Select the pressure discretization format as Coupled. Set the convergence criterion to 10 -7 The buoyancy center of the underwater vehicle is defined as the point of action of its moment. The hydrodynamic force and hydrodynamic moment of the underwater vehicle are recorded every 50 iterations and exported as a .txt file. After initializing the fluid domain, the iteration step is set to 800. After the iteration is completed, the .out file is exported to record the forces and moments in all directions of the underwater vehicle, which are recorded as the hydrodynamic data in the underwater vehicle velocity coordinate system.

[0031] S3. Establish evaluation indexes for underwater vehicle motion characteristics: The system cmd command and the above-mentioned batch script file are integrated through the iSight platform or other methods, and the optimized Latin hypercube method is used for sampling, which can generate relatively uniform sample points in the entire sample space and has high flexibility. In this embodiment, the number of sample points in the numerical simulation test of the towing tank is 3000, and the number of sample points in the two groups of numerical simulation tests of the spiral arm tank is 900. The main rotation directions of the two groups of tests are pitch direction and yaw direction, respectively. According to the .txt file recording the iteration of hydrodynamic force and hydrodynamic moment, the sample data with a change amplitude greater than 15% in the last 20% iteration step is deleted, and the remaining data constitutes the data set of hydrodynamic force and hydrodynamic moment of the underwater vehicle in each direction.

[0032] S4. Sampling analysis to obtain shape parameters and their hydrodynamic response: The Bayesian regularized back propagation neural network method was used to select the model training method with the highest accuracy. The data set was then evenly divided into 5 groups, 4 of which were used as training sets and the remaining set of data as a validation set. The 5 proxy models were trained separately, and the data with poor prediction results in the 5 proxy models were deleted and the proxy models were retrained. According to the numerical simulation results of the towing tank, the resistanceF D , lift F L , lateral force F SF , rolling moment M rol , yaw moment M yaw and pitch moment M pit The corresponding dimensionless coefficients and the proxy model of the angle of attack and sideslip angle are established based on the numerical simulation results of the swing arm pool. F D , lift F L , lateral force F SF , rolling moment M rol , yaw moment M yaw and pitch moment M pit The corresponding dimensionless coefficients are proxy models for the part related to the angular velocity of rotation.

[0033] S5. Use the data set to build a high-precision proxy model and analyze the sensitivity relationship between shape parameters and motion characteristics: Based on the above proxy model, combined with the interpolation method, the position derivative and rotation derivative of the hydrodynamic coefficient of the underwater vehicle shape under each shape parameter can be obtained. Therefore, the input-response relationship between the shape parameters of the underwater vehicle and its motion characteristics can be established. The Sobol' sensitivity analysis method is used to determine the influence of each parameter on the output, and the variables with less influence can be set as constants during the design process.

[0034] S6. Solve the optimization problem with constraints through the optimization algorithm and obtain the optimal geometric parameters of the shape: According to the established proxy model and motion characteristic model, the shape of the underwater vehicle can be optimized in combination with different mission requirements. In this embodiment, the volume of the middle part of the fuselage of the underwater vehicle remains unchanged, the length from the front to the rear remains unchanged, and the diameter of the body remains unchanged. d ∈[230, 250] mm, the remaining geometric parameters are in the sample space of the data set used by the proxy model, and the diving depth of the underwater vehicle is set to 1000 m. Both the horizontal and vertical plane stability are kept dynamically stable, and four different optimization conditions are applied to the embodiment: Condition 1: When the inn load is 1 W, the endurance economy is as high as possible; Condition 2: Have the greatest possible speed; Condition 3: Have the highest possible turning maneuverability; Condition 4: When the net buoyancy is 5 N, the pitch angle is -10°, and the roll angle is 20°, the turning radius is no greater than 10 m. At the same time, the cruising economy is as high as possible when the inn load is 5 W.

[0035] The above optimization results were verified using computational fluid dynamics methods, and the results are shown in Table 2-5 and Figure 4-7 shown.

[0036] Table 2. Comparison of results before and after optimization of condition 1

[0037] Table 3. Comparison of results before and after optimization of condition 2

[0038] Table 4. Comparison of results before and after optimization of condition 3

[0039] Table 5 Comparison of results before and after optimization of condition 4

[0040] The detailed operation steps and design results of the above-mentioned embodiments show that the rapid design method of the pedigree shape of underwater vehicles has established a parametric model of underwater vehicles, a hydrodynamic numerical simulation analysis method of the shape, an evaluation index of the motion characteristics of underwater vehicles, sampling and analysis to obtain shape parameters and their hydrodynamic response, and a high-precision hydrodynamic proxy model is established using the data set. The optimal geometric parameters of the underwater vehicle shape under different constraints can be quickly solved through the optimization algorithm, which improves the design efficiency and accuracy of the underwater vehicle shape.

[0041] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A rapid design method for a pedigree shape of an underwater vehicle, comprising the following steps: S1. Establish a parameterized model of underwater vehicle; S2. Establish a hydrodynamic numerical simulation analysis model for the shape, including shape meshing and solver settings; S3. Establish evaluation indexes for underwater vehicle motion characteristics; S4. Sampling and analysis to obtain shape parameters and their hydrodynamic response; S5. Use the data set to build a high-precision proxy model and analyze the sensitivity relationship between shape parameters and motion characteristics; S6. Solve the optimization problem with constraints through optimization algorithm to obtain the optimal geometric parameters of the shape.

2. The rapid design method of underwater vehicle genealogy appearance according to claim 1 is characterized in that: The step S1 includes: determining the basic form of the underwater vehicle according to the design requirements of the shape, and establishing an initial three-dimensional shape model using Catia software.

3. The rapid design method of underwater vehicle genealogy appearance according to claim 2 is characterized in that: The step S2 comprises the following steps: importing the three-dimensional model into the ICEM software, setting the angle of attack and sideslip angle of the underwater vehicle and the fluid domain in which it is located, and then performing grid division; importing the obtained grid file into the Fluent software, selecting the pressure-based solver, selecting the k-ω SST turbulence model, setting the boundary type and the incoming flow velocity boundary condition, and selecting the pressure discrete format as Coupled; and performing iterative solution calculation after initialization.

4. The rapid design method of underwater vehicle genealogy appearance according to claim 3 is characterized in that: The step S3 includes: establishing an underwater vehicle dynamics model based on the Newton-Euler method, and analyzing the stability, navigation economy, navigation speed and turning maneuverability of the underwater vehicle dynamics model.

5. The rapid design method of underwater vehicle genealogy appearance according to claim 4 is characterized in that: The step S4 includes the following steps: integrating the steps S1 and S2, determining the shape parameters to be optimized in step S1, taking the angle of attack and sideslip angle in step S2 as input variables, and taking the resistance of the underwater vehicle as input variables. F D , lift F L , lateral force F SF , rolling moment M rol , yaw moment M yaw and pitch moment M pit is the output variable; the sampling test is designed using the optimized Latin hypercube method, and the output response of the sample is calculated using the hydrodynamic numerical analysis method.

6. The rapid design method of underwater vehicle genealogy appearance according to claim 5 is characterized in that: The step S5 includes the following steps: using the above data set, using the cross-validation method, comparing the accuracy of different proxy model training methods to establish proxy models of various hydrodynamics, and selecting the model training method with the highest accuracy. Then the data set is evenly divided into 5 groups, 4 of which are training sets, and the remaining group of data is used as a validation set. The 5 groups of proxy models are trained separately, and the data with poor prediction results in the 5 groups of proxy models are deleted, and the proxy models are retrained. Using the motion characteristic evaluation index established in step S2, the sensitivity of each design variable is analyzed using the Sobol' sensitivity analysis method, which is used to optimize the selection of variables before optimization. Variables with lower sensitivity can be adjusted according to the needs of other design units.

7. The rapid design method of underwater vehicle genealogy appearance according to claim 6 is characterized in that: The step S6 comprises the following steps: according to actual design requirements, comprehensively considering the volume, stability, navigation economy, navigation speed and turning maneuverability of the underwater vehicle, setting constraints and optimization goals, and finding the optimal size design scheme of the appearance under specific working conditions.

Images