A numerical simulation method for flow field of autonomous swimming underwater bionic fish

Abstract

The present invention discloses a method for numerically simulating the flow field of an underwater bionic fish swimming autonomously, and specifically relates to the field of numerical simulation of the flow field of a bionic fish swimming, including: describing the shape of the fish body by establishing a contour line equation of a horizontal section. The global ground coordinate system is associated with the fish body coordinate system through the center of gravity position and pitch angle of the fish body, and a time-series fish body motion data set is established. The time-series flow field data set is generated by combining the non-steady-state flow field data collected by a distributed underwater sensor array. Based on the flow field control equation and the dynamic motion characteristics of the fish body, a static interactive coupling model is constructed, and key dynamic characteristic quantities are extracted to establish a flow field prediction model. The interactive coupling data is input into the static model to calculate the global coupling solution for error analysis, and a dynamic interactive coupling model is generated. According to the flow field prediction results and the dynamic interactive coupling model, the motion characteristic data of the fish body is corrected, the motion trajectory of the bionic fish is optimized, and the closed-loop optimization of the flow field and the fish body motion is realized.

Description

Technical Field

[0001] The invention relates to the technical field of numerical simulation of flow field of bionic fish swimming, and more specifically, to a numerical simulation method of flow field of autonomous swimming of underwater bionic fish. Background Art

[0002] In the field of underwater motion simulation, especially in the motion simulation of bionic fish, the interaction between fluid and fish plays a vital role. Traditional underwater motion simulation usually relies on steady-state flow field models or simplified fish motion equations, but in practical applications, due to the non-steady-state nature of water flow and the complex motion of bionic fish, these methods often have difficulty in accurately capturing the dynamic coupling relationship between flow field and fish motion.

[0003] Therefore, how to achieve accurate fish motion prediction through dynamic modeling of the complex interaction between fluid and fish in a complex flow field environment, and then regulate the motion parameters based on the prediction results has become an important technical challenge.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a flow field numerical simulation method for autonomous swimming of an underwater bionic fish to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] S1: Divide the bionic fish body into the head, trunk and tail regions along the horizontal direction, and establish the contour line equation of the fish body shape in the horizontal section;

[0008] S2: The global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle to establish a time series fish body motion characteristic data set;

[0009] S3: Use a distributed underwater sensor array to collect unsteady flow field data in the target waters, extract velocity vectors, vortex structures, and pressure distribution changes, and establish a time-series flow field data set;

[0010] S4: Based on the control equations of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, a static interactive coupling model of the flow field effect and the fish body motion characteristic response is constructed;

[0011] S5: Extract key dynamic characteristics including velocity gradient, pressure distribution change and local vortex structure from the time series flow field data set, and establish a flow field prediction model;

[0012] S6: Input the interactive coupling data based on the time series into the static interactive coupling model to calculate the global coupling solution, perform error analysis on the global coupling solution and the flow field prediction model, and form a dynamic interactive coupling model;

[0013] S7: Predict the flow field change according to the flow field prediction model, and then calculate the predicted motion characteristic data of the fish body in combination with the dynamic interactive coupling model to correct the deviation of the motion characteristic data.

[0014] In a preferred embodiment, in S1, the bionic fish body is divided into the head, trunk and tail regions along the horizontal direction, and the contour line equation of the fish body shape in the horizontal section is established as follows:

[0015] ;

[0016] In the formula, is the arc length parameter of the fish body midline, the distance from the head to the tail along the midline, ranging from 0 to , is the width of the fish body in horizontal section, , is the maximum width of the head and the width of the tail, is the dividing point between the head and torso. is the dividing point between the trunk and the tail. is the total length of the fish.

[0017] In a preferred embodiment, in S2, the global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle, and the establishment of the time series fish body motion characteristic data set specifically includes:

[0018] Establish a global ground coordinate system and a fish body coordinate system. The origin of the fish body coordinate system is set at the center of gravity of the fish body. The direction of the coordinate axis is defined by the line from the head of the fish body to the center of gravity and is perpendicular to the shape of the fish body.

[0019] The background Cartesian grid is defined in the global ground coordinate system, and the Lagrangian grid of the fish body boundary is defined in the fish body coordinate system;

[0020] The fish body coordinate system is translated and rotated according to the autonomous swimming of the fish body. The fish body pitch angle is used to associate the two coordinate systems, and the coordinate conversion is completed through the center of gravity position and pitch angle;

[0021] The real-time position change trajectory of the fish body in the global ground coordinate system, the force characteristics of the fish body, and the tail swing characteristics are monitored and recorded in real time to establish a time-series fish body motion characteristic data set.

[0022] In a preferred embodiment, in S3, using a distributed underwater sensor array to collect unsteady flow field data in the target water area, extracting velocity vectors, vortex structures and pressure distribution changes, and establishing a time series flow field data set specifically includes:

[0023] Distributed fluid sensing equipment is arranged within the range of the bionic fish in the global ground coordinate system, including a flow velocity sensing device, a pressure measuring device and a vortex sensing device;

[0024] The distributed fluid sensing device captures the velocity vector parameters, pressure change parameters and vortex characteristic parameters in the target water area at a preset sampling frequency, and marks the above parameters as the fluid raw data stream;

[0025] The raw fluid data streams collected by fluid sensing devices at different spatial positions in the global ground coordinate system are integrated and processed according to the preset sampling frequency. The integrated data are dynamically filtered for noise and interference data through frequency domain analysis technology, and the remaining data are converted into a time-series flow field data set.

[0026] In a preferred embodiment, in S4, based on the control equation of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, constructing a static interactive coupling model of the flow field effect and the fish body motion characteristic response specifically includes:

[0027] Based on the continuity and momentum conservation laws of unsteady flow and the predefined boundary conditions of the fish movement area, the control equations describing the unsteady flow field characteristics of the dynamic changes of the velocity field and pressure field around the fish are constructed;

[0028] The flexible structure segmented fitting method is used, combined with the contour line equation of the horizontal section, to define the mathematical comprehensive expression of the dynamic motion characteristics of different parts of the fish body;

[0029] By establishing the relationship between the force on the fish surface and the expression of its motion characteristics, the flow field effect and the response of the fish's motion characteristics are combined into a static interactive coupling model.

[0030] In a preferred embodiment, in S5, extracting key dynamic feature quantities including velocity gradient, pressure distribution change and local vortex structure from the time series flow field data set, and establishing a flow field prediction model specifically includes:

[0031] Perform multi-dimensional degradation operations on the time-series flow field data set to extract key dynamic feature quantities including velocity gradient, pressure distribution changes, and local vortex structure to form a set of flow field characterization parameters;

[0032] Using characteristic mode decomposition technology, the flow field characterization parameters are divided into spatial and temporal scale components. By independently analyzing the local dynamic behavior and global change trend of the flow field characterization parameters at different scales, the flow field motion is hierarchically modeled.

[0033] The flow field historical data in the time series flow field data set is used as the flow field characterization parameter input to capture the dynamic evolution trend of the flow field and build a flow field prediction model composed of a recursive neural network.

[0034] The flow field evolution results generated by the flow field prediction model are compared with the actual time series data for error analysis, and the flow field prediction model is corrected by adjusting the model parameters.

[0035] In a preferred embodiment, in S6, the interactive coupling data based on the time series is input into the static interactive coupling model to calculate the global coupling solution, and the error analysis between the global coupling solution and the flow field prediction model is performed to form the dynamic interactive coupling model, which specifically includes:

[0036] Based on the mapping principle of time series alignment and global ground coordinate system spatial distribution, the data of the time series flow field data set and the time series fish body motion characteristic data set are unified and integrated to form interactive coupling data based on time series, and the interactive coupling data are input into the static interactive coupling model;

[0037] The mathematical comprehensive expression of the control equations of the unsteady flow field characteristics and the fish body motion characteristics is transformed into a discretized algebraic equation system, and the discrete form includes a piecewise representation of the space and time steps;

[0038] The discretized algebraic equations are dynamically distributed to different processing units of the graphics processor architecture for distributed iterative operations, the algebraic equations of each processing unit are solved independently, and the iteration rules are set to adjust the local solutions of the processing units to achieve the convergence of the global coupled solution;

[0039] Extract the flow field prediction data and the fish body motion characteristic data from the flow field prediction model and the global coupling solution, compare them with the flow field data and the fish body motion characteristic data actually collected at the corresponding prediction time point, and generate a parameter error matrix;

[0040] Based on the error analysis of the parameter error matrix, the predefined boundary conditions of the fish movement area in the static interactive coupling model and the model parameters of the flow field prediction model are adjusted;

[0041] Based on the adjusted boundary conditions and the prediction results of the flow field prediction model, the global dynamic coupling solution is iteratively calculated, the error is repeatedly analyzed and the model parameters of the predefined boundary conditions and the flow field prediction model are adjusted until the error of the global coupling solution converges to the set acceptable range, thus forming a dynamic interactive coupling model.

[0042] In a preferred embodiment, in S7, the flow field change is predicted according to the flow field prediction model, and the predicted motion characteristic data of the fish body is calculated in combination with the dynamic interactive coupling model, and the deviation of the motion characteristic data is corrected, which specifically includes:

[0043] Analyze the real-time collected flow field data, use the velocity distribution, pressure gradient and vortex position in the form of time and space distribution as the flow field prediction input, and predict the flow field changes according to the flow field prediction model;

[0044] The target motion characteristic data is set, and the predicted motion characteristic data of the fish body is calculated based on the flow field data prediction results and combined with the dynamic interactive coupling model. According to the deviation between the target motion characteristic and the predicted motion characteristic data, the tail swing parameters of the fish body are dynamically adjusted to correct the motion characteristic data deviation until the predicted motion characteristic data matches the set target motion characteristic data.

[0045] The technical effects and advantages of the flow field numerical simulation method of the autonomous swimming of underwater bionic fish of the present invention are as follows:

[0046] This scheme accurately captures the complex interaction between the underwater flow field and the bionic fish body through a multi-scale dynamic simulation method based on real-time flow field reconstruction, and has significant technical effects and advantages. First, by reconstructing the underwater flow field in real time and combining the motion characteristics of the fish body, the scheme can dynamically predict the changes in the flow field and reflect the impact of the flow field on the motion of the bionic fish body, thereby providing a more accurate motion simulation. This dynamic adaptability based on flow field changes ensures high-precision motion prediction of the bionic fish in different flow environments. Secondly, combined with the dynamic feature extraction method, it can capture the non-steady-state changes of the water flow in real time, overcome the limitations of the traditional static flow field model, and improve the real-time and accuracy of the model in the dynamic water environment. In addition, through the dynamic optimization mechanism of the static interactive coupling model, the effect of the flow field and the motion characteristics of the fish body can respond synchronously, ensuring higher physical consistency and simulation accuracy, and providing a reference basis for early correction of the deviation of the motion parameters of the bionic fish in the non-steady-state flow field. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of a flow field numerical simulation method of an underwater bionic fish swimming autonomously according to the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Embodiment 1, Figure 1 The present invention provides a flow field numerical simulation method for autonomous swimming of an underwater bionic fish, which comprises the following steps:

[0050] S1: Divide the bionic fish body into the head, trunk and tail regions along the horizontal direction, and establish the contour line equation of the fish body shape in the horizontal section;

[0051] S2: The global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle to establish a time series fish body motion characteristic data set;

[0052] S3: Use a distributed underwater sensor array to collect unsteady flow field data in the target waters, extract velocity vectors, vortex structures, and pressure distribution changes, and establish a time-series flow field data set;

[0053] S4: Based on the control equations of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, a static interactive coupling model of the flow field effect and the fish body motion characteristic response is constructed;

[0054] S5: Extract key dynamic characteristics including velocity gradient, pressure distribution change and local vortex structure from the time series flow field data set, and establish a flow field prediction model;

[0055] S6: Input the interactive coupling data based on the time series into the static interactive coupling model to calculate the global coupling solution, perform error analysis on the global coupling solution and the flow field prediction model, and form a dynamic interactive coupling model;

[0056] S7: Predict the flow field change according to the flow field prediction model, and then calculate the predicted motion characteristic data of the fish body in combination with the dynamic interactive coupling model to correct the deviation of the motion characteristic data.

[0057] In S1, the bionic fish body is divided into the head, trunk and tail regions along the horizontal direction, and the contour line equation of the fish body shape in the horizontal section is established as follows:

[0058] ;

[0059] In the formula, is the arc length parameter of the fish body midline, the distance from the head to the tail along the midline, ranging from 0 to , is the width of the fish body in horizontal section, , is the maximum width of the head and the width of the tail, is the dividing point between the head and torso. is the dividing point between the trunk and the tail. is the total length of the fish.

[0060] In S2, the global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle. The establishment of the time series fish body motion characteristic data set specifically includes:

[0061] The global ground coordinate system is set as a Cartesian coordinate system, with its origin at a predetermined reference point in the water area where the bionic fish is located. The axis direction of the global coordinate system is consistent with the coordinate system in the water area. The fish body coordinate system is set as a local coordinate system relative to the center of gravity of the fish body, where the center of gravity of the fish body is the coordinate origin. The coordinate axis direction is defined by the line from the head of the fish body to the center of gravity, and is perpendicular to the shape of the fish body, so as to clearly describe the translation and rotation of the fish body.

[0062] The background Cartesian grid in the global ground coordinate system is used to describe the overall flow field of the water area. The grid covers the target water area and is updated in real time as the water flow changes dynamically. At the same time, the Lagrangian grid in the fish coordinate system is used to describe the boundary conditions of the fish body, and the grid points always follow the movement of the fish body. These grids ensure that the interaction between the water flow and the fish body can be accurately captured at local and global scales.

[0063] The variables in the fish body coordinate system need to be linked to the global ground coordinate system through the fish body center of gravity position and attack angle. Assume that at a certain moment, the coordinates of the bionic fish's center of gravity in the global ground coordinate system are , the angle of attack is , then any point on the fish body coordinate system The coordinates in the global ground coordinate system are:

[0064] ;

[0065] In the formula, yes The coordinates in the global ground coordinate system after transformation, They are all time-related functions and can be obtained by solving the motion control equation of the bionic fish. The position of the center of gravity coordinates in the fish body coordinate system is:

[0066] ;

[0067] In the formula, is the position of the center of gravity coordinate in the fish body coordinate system, For the quality of the bionic fish, is the coordinate of the center of gravity of any mass unit of the fish body in the fish body coordinate system, is the density of the mass unit. The conversion between the two coordinate systems is achieved by calculating the relationship between the fish's center of mass and the pitch angle.

[0068] The position changes of the fish body in the global ground coordinate system are captured in real time, including the trajectory of the fish body in the water, the force characteristics of the fish body, and the tail swing characteristics. The force characteristics include lift, drag and thrust, and the tail swing characteristics include swing amplitude and frequency. At the same time, the velocity vector and acceleration data are calculated through the motion trajectory. To this end, a real-time sensor array is used to collect the force data on the surface of the fish body, and combined with the angle information of the tail swing, the movement process of the fish body is accurately described. These data are recorded in a time series manner to generate a comprehensive fish body motion characteristic data set.

[0069] In S3, a distributed underwater sensor array is used to collect the unsteady flow field data of the target waters, extract the velocity vector, vortex structure and pressure distribution changes, and establish a time series flow field data set, which includes:

[0070] Distributed fluid sensing devices, including flow velocity sensing devices, pressure measuring devices and vortex sensing devices, are deployed within the range of the bionic fish in the global ground coordinate system. Each device is distributed in different spatial positions according to the characteristics of the target water area, covering the area where the bionic fish may pass, so as to monitor the fluid conditions in the water area in real time. These devices operate at a preset sampling frequency to ensure that various parameters of the water flow are obtained in real time.

[0071] Distributed fluid sensing equipment captures the flow velocity vector parameters, pressure change parameters and vortex characteristic parameters in the target water area at a preset sampling frequency, and marks the above parameters as fluid raw data streams. These data streams come from different spatial locations, forming an extensive flow field monitoring network, providing basic data for subsequent flow field reconstruction.

[0072] The raw data streams of fluid collected by fluid sensing devices at different spatial positions in the global ground coordinate system are integrated and processed according to the sampling frequency. The integrated data is dynamically filtered for noise to eliminate interference data through frequency domain analysis technology. Frequency domain analysis can effectively identify high-frequency noise and irregular fluctuations, and maintain the stability and accuracy of the data. The data after noise filtering retains the flow field change trend in the target waters. The remaining data is converted into a time-series flow field data set.

[0073] In S4, based on the control equations of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, a static interactive coupling model of the flow field effect and the fish body motion characteristic response is constructed, which specifically includes:

[0074] According to the continuity and momentum conservation law of unsteady flow, combined with the predefined boundary conditions of the fish movement area, the control equation describing the unsteady flow field characteristics of the dynamic changes of the velocity field and pressure field around the fish is constructed. The control equation is:

[0075] ;

[0076] In the formula, is the density of the fluid, represents the time rate of change of the fluid velocity field, is the velocity field, is the nonlinear convection term, describing the interaction between fluid particles, is the pressure gradient, is the dynamic viscosity of the fluid, is the viscous diffusion term of the fluid, and F is the external force term, which represents the interaction force between the fish body and the fluid (here it represents the pressure corresponding to the pressure field).

[0077] The fish body motion response equation is:

[0078] ;

[0079] In the formula, is the resultant force on the fish surface, p is the pressure on the fish surface, is the unit normal vector of the fish surface, is the area of ​​the surface element, represents the velocity gradient tensor of the fluid, which represents the spatial variation of the fluid velocity field. , , are the mass matrix, damping matrix, and stiffness matrix of the fish body, respectively. q is the global ground coordinate system coordinate of the fish body. , Represent the acceleration and velocity of the fish body respectively.

[0080] A flexible structure segmented fitting method is adopted (polynomial fitting of different orders is selected based on the complexity of segmented motion. In order to reduce the computational complexity, this scheme adopts quadratic polynomial) and the contour line equation of the horizontal section is combined to define the mathematical comprehensive expression of the dynamic motion characteristics of different parts of the fish body.

[0081] By combining the above equations, a static interactive coupling model describing the flow field and the motion characteristic response of the fish body is obtained. This model can accurately describe the relationship between the movement of the fish body and the change of the flow field under the interaction between the fish body and the fluid, providing a theoretical basis for further dynamic analysis and optimization.

[0082] In S5, key dynamic characteristics including velocity gradient, pressure distribution change and local vortex structure are extracted from the time series flow field data set, and the flow field prediction model is established, which specifically includes:

[0083] The time-series flow field data set collected from the distributed sensor array contains data in multiple spatial and temporal dimensions, such as velocity vectors, pressure changes, and vortex structures. The purpose of multi-dimensional degradation of these high-dimensional data is to extract the feature quantities that best reflect the dynamic evolution of the flow field. Using dimensionality reduction algorithms, such as principal component analysis (PCA) or flow field feature extraction algorithms, velocity gradients, pressure distribution changes, and local vortex structures are extracted as key features of the flow field to form a set of flow field characterization parameters. These feature quantities can not only effectively reduce the data dimension, but also retain the key dynamic information of the flow field evolution.

[0084] By applying characteristic mode decomposition (EMD) to the above set of flow field characterization parameters, we can decompose the time series data of the flow field into different modal components. Each mode represents a scale component in the dynamic behavior of the flow field. Specifically, the EMD technology is used to divide the flow field characterization parameters into spatial scale and time scale components, and the local dynamic behavior and global change trends of these components are analyzed independently. For example, the dynamic changes of short-term local vortices may be significantly different from the overall trend of the flow field on a long time scale. Processing these components separately helps to improve the accuracy of flow field modeling.

[0085] Based on the extracted flow field characterization parameters and the decomposed spatial and temporal scale components, a recurrent neural network (RNN) is used to build a flow field prediction model. RNN can effectively capture the dynamic change rules in time series data, so it is suitable for predicting the dynamic evolution trend of the flow field. In the input layer, historical data in the time series flow field data set is used as input, including historical records of characteristics such as velocity, pressure, and vortex; in the output layer, the recurrent neural network outputs the future evolution trend of the flow field and predicts the flow field behavior in the future. The model can capture the time series pattern of the flow field from historical data and achieve efficient flow field prediction.

[0086] After the prediction is completed, the flow field evolution results generated by the model are compared with the actual collected time series data to calculate the error. Error analysis can reflect the accuracy and adaptability of the model by evaluating the difference between the model prediction results and the actual observation results. If the error exceeds the predetermined threshold, the model needs to be corrected. This correction process usually adjusts the parameters in the flow field prediction model, such as the learning rate of the recursive network, the number of hidden layer neurons, the activation function, etc., to improve the prediction ability of the model under specific flow field conditions. By repeatedly iterating this process, the optimized flow field prediction model can provide more accurate predictions of flow field evolution trends.

[0087] In S6, the interactive coupling data based on the time series is input into the static interactive coupling model to calculate the global coupling solution, and the error analysis between the global coupling solution and the flow field prediction model is performed to form a dynamic interactive coupling model, which specifically includes:

[0088] Based on the mapping principle of time series alignment and spatial distribution of the global ground coordinate system, the data of the time series flow field dataset and the time series fish motion characteristic dataset are unified and integrated to form interactive coupling data based on time series, and the interactive coupling data are input into the static interactive coupling model.

[0089] The control equations of the unsteady flow field characteristics and the mathematical comprehensive expression of the fish body motion characteristics are transformed into a discretized algebraic equation system, which includes a segmented representation of the space and time steps to ensure the continuity and accuracy of the physical quantity calculation at each time step. For example, the equations of the velocity field and pressure field are discretized at each grid point, so that these physical equations can be effectively solved in numerical calculations.

[0090] According to the discretized algebraic equations, they are dynamically allocated to multiple processing units of the graphics processing unit architecture (GPU) for parallel computing. Each processing unit is responsible for solving the local algebraic equations and adjusting the local solution of the processing unit according to the set iteration rules, so that the local solution can be coupled with the global solution to ensure convergence to the global coupled solution. In this process, each processing unit not only independently calculates the solution of its corresponding area, but also transmits the boundary conditions of the solution to the adjacent area through the parallel communication mechanism, thereby achieving accurate convergence of the global coupled solution.

[0091] Extract the flow field prediction data and the fish body motion characteristic data from the flow field prediction model and the global coupling solution, compare them with the flow field data and the fish body motion characteristic data actually collected at the corresponding prediction time point, and generate a parameter error matrix;

[0092] According to the analysis results of the parameter error matrix, the predefined boundary conditions of the fish movement area in the static interactive coupling model and the model parameters of the flow field prediction model are adjusted. The adjustment of boundary conditions may involve the interaction between the fish surface and the flow field, such as friction coefficient, fish movement range, etc.; the adjustment of model parameters includes physical parameters in the flow field prediction model, such as flow velocity, vortex intensity, etc. Through this adjustment, the model can more accurately reflect the coupling effect between the actual flow field and the fish movement, thereby improving the prediction accuracy.

[0093] Based on the adjusted boundary conditions and flow field prediction model, the global dynamic coupling solution is calculated again. By inputting the adjusted parameters into the static interactive coupling model, repeated iterative operations are performed until the error of the global coupling solution converges to the set acceptable range. In this process, the system continuously performs error analysis, analyzes the difference between the flow field prediction and the actual data, gradually corrects the model, and optimizes the prediction accuracy. Finally, an optimized dynamic interactive coupling model is obtained, which can accurately predict the interaction behavior between fish movement and flow field.

[0094] In S7, the flow field change is predicted according to the flow field prediction model, and the predicted motion characteristic data of the fish body is calculated in combination with the dynamic interaction coupling model, and the deviation of the motion characteristic data is corrected, specifically including:

[0095] The flow field prediction model combines the parameters obtained through training with the real-time collected flow field data to generate a series of flow field prediction results at future time points, including changes in velocity field, pressure gradient and vortex structure. This prediction result can not only provide the future trend of flow field evolution, but also provide key information for the prediction of bionic fish body motion characteristics.

[0096] The target motion characteristic data (such as the ideal swimming trajectory of the fish in the water, the tail swing angle, etc.) are set, combined with the flow field prediction results, and input into the dynamic interactive coupling model to calculate the motion characteristics of the bionic fish. The dynamic interactive coupling model takes into account the influence of the flow field on the movement of the fish, and obtains the predicted motion characteristic data of the fish through the coupled calculation process. These data include the movement trajectory of the fish, the tail swing angle, the swimming speed, etc., simulating the actual movement of the fish under the action of the flow field.

[0097] By comparing the deviation between the predicted motion characteristic data and the set target motion characteristic data, error analysis is performed. If there is a gap between the predicted motion characteristics and the target set value, the tail swing parameters of the fish body are dynamically adjusted based on these deviations. The adjustment of the tail swing parameters optimizes the motion trajectory of the fish body by controlling the amplitude, frequency and phase of the tail fin swing, so that the motion characteristics gradually approach the target set value. Specifically, by adjusting the dynamic parameters of the tail swing of the bionic fish body, the swimming speed, angle, etc. of the fish body can be corrected, thereby achieving precise motion control. If it is found in the actual calculation that there is an error between the movement of the fish body and the target movement (for example, the deviation of the tail swing angle is large), according to the error correction model, the tail swing parameters are dynamically adjusted to optimize the swimming trajectory and speed of the fish body. This adjustment process will be iterated at each time step, and this process is dynamic until the movement of the fish body is completely consistent with the set target or as consistent as possible. Finally, under the action of the flow field, the movement of the bionic fish can efficiently and accurately reach the target path, optimize the propulsion efficiency and reduce energy consumption.

[0098] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0099] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or may be transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0100] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0101] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0102] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0103] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0104] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0105] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0106] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0107] Finally: 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, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A numerical simulation method for flow field of autonomous swimming underwater bionic fish, characterized in that: The steps include: S1: Divide the bionic fish body into the head, trunk and tail regions along the horizontal direction, and establish the contour line equation of the fish body shape in the horizontal section; S2: The global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle to establish a time series fish body motion characteristic data set; S3: Use a distributed underwater sensor array to collect unsteady flow field data in the target waters, extract velocity vectors, vortex structures, and pressure distribution changes, and establish a time-series flow field data set; S4: Based on the control equations of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, a static interactive coupling model of the flow field effect and the fish body motion characteristic response is constructed; S5: Extract key dynamic characteristics including velocity gradient, pressure distribution change and local vortex structure from the time series flow field data set, and establish a flow field prediction model; S6: Input the interactive coupling data based on the time series into the static interactive coupling model to calculate the global coupling solution, perform error analysis on the global coupling solution and the flow field prediction model, and form a dynamic interactive coupling model; S7: predicting the flow field change according to the flow field prediction model, and then calculating the predicted motion characteristic data of the fish body in combination with the dynamic interactive coupling model, and correcting the deviation of the motion characteristic data; In S5, key dynamic characteristics including velocity gradient, pressure distribution change and local vortex structure are extracted from the time series flow field data set, and the flow field prediction model is established, which specifically includes: Perform multi-dimensional degradation operations on the time-series flow field data set to extract key dynamic feature quantities including velocity gradient, pressure distribution changes, and local vortex structure to form a set of flow field characterization parameters; Using characteristic mode decomposition technology, the flow field characterization parameters are divided into spatial and temporal scale components. By independently analyzing the local dynamic behavior and global change trend of the flow field characterization parameters at different scales, the flow field motion is hierarchically modeled. Taking the flow field historical data in the time series flow field data set as input, the prediction results representing the dynamic evolution trend of the flow field are output, the time series pattern of the flow field is captured, and a flow field prediction model composed of a recursive neural network is constructed; Compare the flow field evolution results generated by the flow field prediction model with the actual time series data for error analysis, and modify the flow field prediction model by adjusting the model parameters; In S6, the interactive coupling data based on the time series is input into the static interactive coupling model to calculate the global coupling solution, and the error analysis between the global coupling solution and the flow field prediction model is performed to form a dynamic interactive coupling model, which specifically includes: Based on the mapping principle of time series alignment and global ground coordinate system spatial distribution, the data of the time series flow field data set and the time series fish body motion characteristic data set are unified and integrated to form interactive coupling data based on time series, and the interactive coupling data are input into the static interactive coupling model; The mathematical comprehensive expression of the control equations of the unsteady flow field characteristics and the fish body motion characteristics is transformed into a set of discretized algebraic equations, and the discrete form includes a piecewise representation of the space and time steps; The discretized algebraic equations are dynamically distributed to different processing units of the graphics processor architecture for distributed iterative operations, the algebraic equations of each processing unit are solved independently, and the iteration rules are set to adjust the local solutions of the processing units to achieve the convergence of the global coupled solution; Extract the flow field prediction data and the fish body motion characteristic data from the flow field prediction model and the global coupling solution, compare them with the flow field data and the fish body motion characteristic data actually collected at the corresponding prediction time point, and generate a parameter error matrix; Based on the error analysis of the parameter error matrix, the predefined boundary conditions of the fish movement area in the static interactive coupling model and the model parameters of the flow field prediction model are adjusted; Based on the adjusted boundary conditions and the prediction results of the flow field prediction model, the global dynamic coupling solution is iteratively calculated, the error is repeatedly analyzed and the model parameters of the predefined boundary conditions and the flow field prediction model are adjusted until the error of the global coupling solution converges to the set acceptable range, thus forming a dynamic interactive coupling model.

2. The method for numerically simulating flow field of an underwater bionic fish swimming autonomously according to claim 1, characterized in that: In S1, the bionic fish body is divided into the head, trunk and tail regions along the horizontal direction, and the contour line equation of the fish body shape in the horizontal section is established as follows: ； In the formula, is the arc length parameter of the fish body midline, the distance from the head to the tail along the midline, ranging from 0 to , is the width of the fish body in horizontal section, , is the maximum width of the head and the width of the tail, is the dividing point between the head and torso. is the dividing point between the trunk and the tail. is the total length of the fish.

3. The method for numerically simulating flow field of autonomous swimming underwater bionic fish according to claim 2, characterized in that: In S2, the global ground coordinate system and the fish body coordinate system are associated through the fish body pitch angle, and the coordinate conversion of the two coordinate systems is completed through the fish body center of gravity position and pitch angle. The establishment of the time series fish body motion characteristic data set specifically includes: Establish a global ground coordinate system and a fish body coordinate system. The origin of the fish body coordinate system is set at the center of gravity of the fish body. The direction of the coordinate axis is defined by the line from the fish head to the center of gravity and is perpendicular to the fish body shape. The background Cartesian grid is defined in the global ground coordinate system, and the Lagrangian grid of the fish body boundary is defined in the fish body coordinate system; The fish body coordinate system is translated and rotated according to the autonomous swimming of the fish body. The fish body pitch angle is used to associate the two coordinate systems, and the coordinate conversion is completed through the center of gravity position and pitch angle; The real-time position change trajectory of the fish body in the global ground coordinate system, the force characteristics of the fish body, and the tail swing characteristics are monitored and recorded in real time to establish a time-series fish body motion characteristic data set.

4. The method for numerically simulating flow field of autonomous swimming underwater bionic fish according to claim 3, characterized in that: In S3, a distributed underwater sensor array is used to collect the unsteady flow field data of the target waters, extract the velocity vector, vortex structure and pressure distribution changes, and establish a time series flow field data set, which includes: Distributed fluid sensing equipment is arranged within the range of the bionic fish in the global ground coordinate system, including a flow velocity sensing device, a pressure measuring device and a vortex sensing device; The distributed fluid sensing device captures the velocity vector parameters, pressure change parameters and vortex characteristic parameters in the target water area at a preset sampling frequency, and marks the above parameters as the fluid raw data stream; The raw fluid data streams collected by fluid sensing devices at different spatial positions in the global ground coordinate system are integrated and processed according to the sampling frequency. The integrated data are dynamically filtered for noise to eliminate interference data through frequency domain analysis technology, and the remaining data are converted into a time-series flow field data set.

5. The method for numerically simulating flow field of autonomous swimming underwater bionic fish according to claim 4, characterized in that: In S4, based on the control equations of the unsteady flow field characteristics and the dynamic motion characteristics of the fish body, a static interactive coupling model of the flow field effect and the fish body motion characteristic response is constructed, which specifically includes: Based on the continuity and momentum conservation laws of unsteady flow and the predefined boundary conditions of the fish movement area, the control equations describing the unsteady flow field characteristics of the dynamic changes of the velocity field and pressure field around the fish are constructed; The flexible structure segmented fitting method is used, combined with the contour line equation of the horizontal section, to define the mathematical comprehensive expression of the dynamic motion characteristics of different parts of the fish body; By establishing the relationship between the force on the fish surface and the expression of its motion characteristics, the flow field effect and the response of the fish's motion characteristics are combined into a static interactive coupling model.

6. The method for numerically simulating flow field of autonomous swimming underwater bionic fish according to claim 5, characterized in that: In S7, the flow field change is predicted according to the flow field prediction model, and the predicted motion characteristic data of the fish body is calculated in combination with the dynamic interaction coupling model, and the deviation of the motion characteristic data is corrected, specifically including: Analyze the real-time collected flow field data, use the velocity distribution, pressure gradient and vortex position in the form of time and space distribution as the flow field prediction input, and predict the flow field changes according to the flow field prediction model; The target motion characteristic data is set, and the predicted motion characteristic data of the fish body is calculated based on the flow field data prediction results and combined with the dynamic interactive coupling model. According to the deviation between the target motion characteristic and the predicted motion characteristic data, the tail swing parameters of the fish body are dynamically adjusted to correct the motion characteristic data deviation until the predicted motion characteristic data matches the set target motion characteristic data.

Images