10 results about "Aerodynamic coefficient" patented technology

The application provides an interpretable aerodynamic coefficient prediction method and system based on dynamic weighting. The method comprises: obtaining a flight state parameter feature dataset composed of a plurality of sample points and a corresponding aerodynamic coefficient label dataset, performing normalization processing on the feature dataset to obtain standardized sample features; training a group of heterogeneous machine learning models in parallel based on the standardized sample features and the label dataset, and obtaining unbiased prediction values of each model for each sample point through cross-validation; for each sample point, calculating and fusing three types of sample-level evaluation weights based on the unbiased prediction values of each model corresponding to the sample point to generate dynamic fusion weights of each model for the sample point; for each sample point, performing weighted summation on the unbiased prediction values of each model for the sample point by using the dynamic fusion weights corresponding to the sample point to obtain a final fusion prediction value of the sample point; and summarizing the final fusion prediction values of all sample points to output an aerodynamic coefficient prediction result and a corresponding interpretability analysis report.

The present application relates to a kind of ground effect wing ship longitudinal coefficient estimation method and its aerodynamic characteristic analysis system, belong to the technical field of ground effect wing ship. It covers preliminary design ground effect wing ship scheme, design and manufacture wind tunnel model, obtain data by 3 times variable flight height test, and utilize these data to deduce the mathematical expression of ground effect wing ship longitudinal aerodynamic coefficient, estimate the longitudinal aerodynamic characteristic in any flight height state by the mathematical expression of ground effect wing ship longitudinal aerodynamic coefficient, to simplify test procedure, reduce test number. The present application significantly reduces the workload of model blowing test by combining theoretical analysis with a small amount of test data, shortens the research and development cycle, reduces the consumption of funds. At the same time, the method can accurately obtain the key partial derivative and relative ground effect height in the stability calculation of ground effect wing ship, improve the design efficiency and accuracy, and has important promoting effect on the development and application of ground effect wing ship technology.

This invention provides a method and related apparatus for assessing leading-edge erosion of wind turbine blades, belonging to the technical field of wind turbine blade reliability assessment. The method includes: acquiring meteorological data of the wind turbine to be assessed; discretizing the wind turbine blade along its spanwise direction into multiple blade element micro-segments; calculating a bidirectional coupling correction factor for each rainfall event recorded at each blade element micro-segment based on the instantaneous surface morphology of the wind turbine blade; calculating the Miner total cumulative damage variable based on this correction factor, and then calculating the aerodynamic coefficient after damage; calculating the aerodynamic torque based on the aerodynamic coefficient, and then calculating the equivalent annual power generation loss rate, ultimately determining the degree of erosion attenuation of the wind turbine. This invention quantifies the bidirectional coupling mechanism of rain and sand by introducing a liquid resistance-solid buffer attenuation factor and a solid-promoting liquid-water hammer amplification factor, and achieves phased determination of the latency period and mass loss period based on Miner linear cumulative damage, enabling efficient and accurate assessment of the time-varying impact of blade leading-edge erosion on annual power generation.

This invention belongs to the field of unmanned aerial vehicle (UAV) control technology, specifically relating to a method and system for attitude stabilization control of UAVs under special environmental disturbances. The method includes the following steps: S1: Establishing a wind field disturbance model, a dust particle impact model, and an electromagnetic interference model, and dynamically coupling these three models into a total disturbance vector experienced by the UAV; S2: Constructing a six-degree-of-freedom dynamic model incorporating the influence of the total disturbance vector. This model describes the translational and rotational motion of the UAV's center of mass under multi-source disturbances in complex environments; S3: Based on traditional aerodynamic models and incorporating environmental disturbance parameters, constructing a variable structure model with aerodynamic coefficients. This invention effectively improves the UAV's adaptability to multi-source coupled disturbances by constructing a unified multi-layered architecture for environmental identification, disturbance observation, Lyapunov robust control, and actuator compensation, thereby ensuring mission accuracy and safety.

A path planning method based on analytical gradients of a neural network includes: acquiring sample datasets of a rocket under multiple typical flight conditions; training the neural network with flight control variables as input and aerodynamic coefficients as output until convergence is met, resulting in a surrogate model for fitting the rocket's aerodynamic characteristics; collecting the rocket's current state vector and control vector, inputting the control vector into the surrogate model to obtain the current aerodynamic coefficients, and constructing the rocket's flight state equation; based on the flight state equation, using an automatic differentiation method, analytically calculating the gradient with respect to the control vector to form a gradient sensitivity matrix; using the gradient sensitivity matrix to determine the descent direction; and iteratively updating the control vector using a numerical optimization algorithm until the termination condition is met, outputting the optimal flight path. The optimal flight path obtained by this method more closely matches the rocket's actual aerodynamic environment, exhibiting higher terminal accuracy.

This invention discloses a transfer learning method for aerodynamic modeling based on symbolic regression, belonging to the field of aerodynamic modeling technology in aerodynamics. The method includes: collecting aerodynamic data of a source domain aircraft under various flow conditions, including flow state variables and aerodynamic coefficients; in the pre-training stage, constructing a correlation function expression for the aerodynamic coefficients of the source domain aircraft based on the flow state variables, combined with symbolic regression and shape parameter optimization methods; collecting sparse sample aerodynamic data of a target domain aircraft; in the fine-tuning stage, fine-tuning the correlation function parameters and performing aerodynamic modeling based on the sparse samples of the target domain, evaluating the optimal parameters of the correlation function, and using the optimal correlation function to perform polynomial fitting on the sparse sample aerodynamic data, thereby achieving transfer learning for aerodynamic modeling based on symbolic regression. This invention solves the problems of large aerodynamic data requirements and unreliable aerodynamic model extrapolation in existing aerodynamic database construction methods.

The application provides a method and a terminal for calculating aerodynamic performance of a fan blade. The method comprises: obtaining a transition point parameter on the fan blade; inputting the transition point parameter as a boundary constraint condition into an aerodynamic simulation model to calculate blade airfoil aerodynamic coefficients in a first angle of attack range; and obtaining blade airfoil aerodynamic performance data through an extrapolation model according to the blade airfoil aerodynamic coefficients in the first angle of attack range. The application uses the transition point as an erosion proxy parameter and combines simulation and extrapolation, so that complete airfoil aerodynamic performance data in a full angle of attack range can be quickly and accurately obtained without constructing a complex erosion geometry.

This application discloses a method for identifying aerodynamic parameters of a tail-saddle unmanned aerial vehicle (UAV) and applying it to a data processing device. The method includes: obtaining a propeller thrust coefficient function; obtaining steady-state flight data collected by airborne sensors when the tail-saddle UAV is in steady-state flight at different pitch and sideslip angles; calculating propeller thrust based on the steady-state flight data and the propeller thrust coefficient function; calculating aerodynamic forces by combining dynamic equations and decomposing them into lift, drag, and sideslip forces; calculating aerodynamic coefficients at different angles of attack and sideslip angles; and obtaining the aerodynamic coefficient function based on a polynomial model and least squares fitting. This application does not require specialized equipment such as wind tunnels, but only relies on low-cost airborne sensors to complete aerodynamic identification covering the entire flight envelope in a real flight environment, reducing the cost of engineering applications.

This invention relates to the field of automotive aerodynamic performance prediction technology, providing a cross-dimensional 2D-3D automotive aerodynamic prediction method based on artificial intelligence algorithms. First, sensitive geometric features from the vehicle's normal perspective are selected for parametric deformation, generating a parametric deformation table and a vehicle model set. Two-dimensional and three-dimensional CFD simulations are then performed. Second, based on the mapping relationship between the parametric deformation table and aerodynamic coefficients, a hybrid K-Means and GMM clustering strategy is used to classify the samples. The optimal number of clusters is determined by combining MAE and MSE indices. For each sample class, a multi-branch neural network model set is constructed and trained. Finally, the deformation parameters of the new three-dimensional vehicle body model to be predicted are input into the model set. After clustering localization and preliminary prediction via branch regression, a cross-dimensional gain coefficient for different scenarios is introduced for correction and compensation, outputting the predicted drag coefficient value. This method can significantly reduce the workload of three-dimensional CFD simulation, ensuring prediction accuracy while significantly saving computational resources.

The application discloses a method for identifying aerodynamic parameters of a UAV based on improved maximum likelihood estimation, and is implemented as follows: an initial aerodynamic coefficient table is generated by using DATCOM, a mathematical expression of the aerodynamic parameters is obtained by using a nonlinear least square polynomial fitting method, and a modeling difficulty of measured data is solved; a modeling process is simplified by combining a confidence interval and aerodynamic physical constraints, and complexity and interpretability of the model are balanced; a high-order coupling term of a polynomial is used to represent system uncertainty, iterative calculation is simplified, a gradient and a Hessian matrix are solved, an LM algorithm is introduced, and problems of covariance singularity, low efficiency and unstable convergence of traditional MLE in processing high-dimensional aerodynamic parameters are solved. The method is improved from three dimensions of aerodynamic model construction, optimization criterion function reconstruction and iterative algorithm optimization, is suitable for identifying aerodynamic parameters of a UAV under high dynamic maneuvering states, and has a 14 times higher calculation efficiency and a relative error of less than 7% compared with traditional MLE.