Runway three-dimensional deformation evaluation method based on undercarriage shimmy enhancement effect
By establishing a detailed simulation model of the entire aircraft and a three-dimensional deformation solid model of the runway, and combining measured data and simulation models, the excitation mechanism of runway three-dimensional deformation on aircraft shimmy was analyzed. This solved the problem of the unrevealed dynamic interaction between runway three-dimensional deformation and landing gear, and improved the accuracy and safety of the analysis.
Patent Information
- Application Number
- CN202410987648.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-12
Smart Images

Figure CN118940574A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of aircraft ground operation safety, and in particular to a runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect. Background Art
[0002] When an aircraft is taxiing on the runway at high speed, a slight deviation of the front wheel may evolve into a violent S-shaped shimmy, causing the landing gear to bend and twist at high frequency and swing sharply, resulting in damage to the landing gear or even loss of control of the aircraft. The three-dimensional deformation of the runway is one of the main causes of aircraft landing gear shimmy. The dynamic interaction between the three-dimensional deformation and the aircraft wheels can not only stimulate shimmy, but also make the shimmy process more violent.
[0003] At present, the analysis of runway-induced shimmy is in its infancy, and has not been able to further reveal the dynamic interaction between the three-dimensional deformation of the runway and the landing gear. At the same time, existing studies that consider runway deformation mostly use a simplified front landing gear model, and the consideration of runway deformation is also limited to longitudinal two-dimensional deformation, ignoring the multi-point dynamic contact between the aircraft and the runway and the lateral correlation of the three-dimensional deformation of the runway, and the analysis accuracy needs to be improved. The analysis of runway-induced shimmy is in its infancy, and has not been able to further reveal the dynamic interaction between the three-dimensional deformation of the runway and the landing gear. Shimmy studies considering runway excitation have proved that runway deformation has an adverse effect on shimmy, but have not further analyzed what form and degree of runway deformation can excite shimmy and what the reinforcing effect on shimmy is during the evolution process, making it difficult to effectively deal with the challenges brought by landing gear shimmy. Summary of the invention
[0004] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a runway three-dimensional deformation evaluation method based on the landing gear shimmy enhancement effect. The method establishes a refined analysis method for aircraft shimmy under runway three-dimensional deformation, grasps the influence of three-dimensional deformation characteristics on shimmy excitation, and clarifies the shimmy excitation conditions of three-dimensional deformation. In this way, the three-dimensional deformation of the runway is controlled to reduce the potential risk and harm of shimmy, and improve the airworthiness safety of the runway.
[0005] The technical solution adopted to achieve the purpose of the present invention is a runway three-dimensional deformation evaluation method based on the landing gear shimmy enhancement effect, the method comprising:
[0006] Establish a detailed simulation model of the entire aircraft;
[0007] Establish a three-dimensional deformable solid model of the runway;
[0008] Establishing a runway-aircraft taxiing dynamics simulation model for shimmy analysis based on the aircraft full-machine refined simulation model and the runway three-dimensional deformation solid model;
[0009] According to the runway-aircraft taxiing dynamic simulation model, an enhancement mechanism of the shimmy evolution process is obtained;
[0010] According to the enhancement mechanism of the shimmy evolution process, a characterization model and an index set for describing the shimmy response increment are established;
[0011] According to the characterization index set of the shimmy response increment, the influence of geometric characteristics on the shimmy response increment is obtained;
[0012] According to the influence of geometric characteristics on the shimmy response increment, a mapping model between the shimmy response increment and the three-dimensional deformation is established. The three-dimensional deformation of the runway is evaluated based on the mapping model.
[0013] The present invention has the following advantages:
[0014] (1) Compared with the conventional method of suppressing aircraft shimmy from the aircraft level, the present invention starts from the runway level and evaluates the three-dimensional deformation of the runway through the excitation mechanism of shimmy.
[0015] (2) Develop a parameter self-optimization method that integrates data-driven and dynamic simulation by combining measured and perceived data with simulation models to improve the accuracy of vibration analysis.
[0016] (3) Based on the time-frequency analysis of non-stationary processes, the response increment of the runway's three-dimensional deformation-induced shimmy evolution process is revealed, and a mapping model between the shimmy response increment and the runway's three-dimensional deformation is constructed. The three-dimensional deformation of the runway is evaluated using this mapping model. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 The present invention is a flow chart of the runway three-dimensional deformation evaluation method based on the landing gear shimmy enhancement effect.
[0018] Figure 2 Schematic diagram of the structure of the three-dimensional deformation of the runway in this embodiment.
[0019] Figure 3 Schematic diagram for establishing a three-dimensional deformable solid model of the runway.
[0020] Figure 4 Schematic diagram comparing vibration responses under continuous excitation with and without three-dimensional deformation.
[0021] Figure 5 It is the mapping relationship between three-dimensional deformable geometric features and response increments. DETAILED DESCRIPTION
[0022] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0023] The runway three-dimensional deformation evaluation method based on the landing gear shimmy enhancement effect of the present invention is as follows Figure 1 As shown, the following steps are included:
[0024] S1. Establish a detailed simulation model of the entire aircraft.
[0025] A detailed simulation model of the entire aircraft is established. This embodiment uses four typical aircraft, namely C919, B737, B757, and B777, as examples for explanation.
[0026] Based on the flexible front landing gear and main landing gear models, the flexible fuselage model is established to assemble the full-aircraft shimmy dynamics model by considering the influence of fuselage motion and elasticity on shimmy. The flexible fuselage uses a finite element model to replace the rigid fuselage. Finally, the aircraft fuselage and landing gear models are imported into the ADMA / Aircraft virtual prototype software platform to assemble the full-aircraft assembly and realize the ground motion behavior simulation of the full-scale aircraft.
[0027] Among them, the key to modeling the flexible landing gear lies in accurately handling the motion relationship of the buffer strut, adding relevant vertical elements such as buffers and shimmy dampers, and selecting an appropriate tire force model. The finite element method is used to establish the strut outer cylinder and piston rod model in HyperMesh, and the linear elastic constitutive relationship is used to divide the HEX grid to improve the calculation efficiency. The motion relationship between the strut outer cylinder and the piston rod ignores the surface contact, and the piston rod vertex is always located on the axis of the outer cylinder, and the lower end of the outer cylinder is always located on the axis of the piston rod. The landing gear of the simulation model adopts a single-cavity variable oil hole buffer with a maximum stroke of 410mm for the front landing gear and 400mm for the main landing gear. Considering the shimmy damper acting on the components of the steering freedom of the landing gear, an anti-shimmy damping torque Md is applied. For shimmy analysis, a complex tire force model is established based on the point contact theory. The tire lateral force Flat is a function of the tire vertical force Fn and the tire side slip angle α.
[0028] S2. Establish a three-dimensional deformable solid model of the runway.
[0029] Based on the measured three-dimensional deformation data of 53 runways and the disease inspection data, this embodiment divides the three-dimensional deformation into three types: geometric mutation type, geometric continuity type and stable random type, such as Figure 2As shown. For non-stationary signals such as geometric mutations and continuous deformations, mutation height Δh, lateral and longitudinal wavelengths wt, and amplitude A are used for characterization; for stationary signals, random spectrum power spectral density PSD and coherence coefficient Coh are used for characterization, and a mathematical characterization model is abstracted. Then, according to the mathematical characterization model, the three-dimensional deformation point cloud data of the runway is generated and input into the MATLAB / Simulink module to generate the three-dimensional coordinates of each node, where the z coordinate is the three-dimensional elevation of the pavement, and the three adjacent nodes are connected into triangles. The adjacent triangles have common edges and are connected to form a three-dimensional solid runway with a length of 2500m and a width of 20m. The geometric parameters and deformation data of the runway model are exported as rdf files, where: the [MODEL] data block defines the type of pavement in the file and the equations used by the solver; the [GLOBAL_PARAMETERS] data block defines parameters such as the pavement geometry; the [DATA_POINTS] data block can input the three-dimensional deformation data of the pavement to generate a physical entity model of the runway as shown in the figure. Figure 3 shown.
[0030] S3. Establish a runway-aircraft taxiing dynamic simulation model for shimmy analysis.
[0031] According to the refined simulation model of the whole aircraft and the three-dimensional deformation solid model of the runway, the runway-aircraft taxiing dynamic simulation model was built through the ADAMS software platform. In order to ensure the accuracy of the simulation model in the three-dimensional deformation swing vibration force analysis, the simulation model was verified as follows:
[0032] Set the initial calculation step length, aircraft taxiing speed and other parameters, output the dynamic response indicators such as the swing vibration frequency and amplitude, and combine the measured airborne perception data to construct the objective function F(θ) as shown in equations (1) to (3):
[0033] F(0)=BV(0)+β,VA,(0)(1)
[0034]
[0035] Where: θ represents the simulation model parameter vector; VF o Indicates frequency difference index; VA o Represents the amplitude difference index; f sr Indicates the simulation calculation frequency; f mr Indicates the measured frequency; a sr Represents the simulation calculation amplitude; a mr represents the measured amplitude, β1 and β2 are coefficients, and ABS is the absolute value.
[0036] Through the integration of dynamic simulation and perception data, a self-optimization algorithm for dynamic parameters is developed. Through the optimal approximation of the objective function, a genetic-particle swarm adaptive optimization algorithm is used to quickly search for the optimal solution of the model dynamic parameters. With the update of perception data, dynamic parameters such as oil damping, anti-sway damping, and landing gear bending stiffness are iteratively calculated to meet the global high matching convergence of simulation results and measured data, and ensure the accuracy of the three-dimensional deformation swing vibration force analysis model.
[0037] S4. Based on the runway-aircraft taxiing dynamic simulation model, the enhancement mechanism of the shimmy evolution process is analyzed. The specific process is as follows:
[0038] After the actual shimmy occurs, it is still continuously excited by the three-dimensional deformation of the runway, causing the vibration response to diverge or slowing down the shimmy convergence process. It is proposed to set up two scenarios with and without continuous excitation of three-dimensional deformation, such as Figure 4 As shown in the figure, a three-dimensional deformation-aircraft full-process dynamic simulation is carried out to obtain the time-history response of the strut torsion angle, lateral bending angle, etc. Then, the power spectrum density diagram of the time domain signal is obtained by using the periodogram method, and the Poincaré section and phase diagram are obtained by using the Poincaré mapping. The enhancement effect of the landing gear vibration response is analyzed from multiple angles, and the divergence and convergence mechanism of the shimmy evolution process under the forced action of the three-dimensional deformation of the runway is clarified, that is, the enhancement mechanism of the shimmy evolution process is obtained.
[0039] S5. Based on the evolutionary process enhancement mechanism, a characterization model and an index set that can accurately characterize the increment of the shimmy response are established. The specific process is as follows:
[0040] On the basis of clarifying the divergence and convergence mechanism of the evolution process, referring to the characterization principles of non-stationary signals in time domain, frequency domain, energy, etc., the incremental characteristics of the support torsion angle, lateral bending angle, and wheel swing angle response in the swing vibration evolution process under different three-dimensional deformation excitations are comprehensively considered. Through collinearity and significance analysis, a characterization model and indicator set that can accurately characterize the swing vibration response increment are established, as shown in Table 1.
[0041]
[0042] Table 1 Characterization of non-stationary signals
[0043] S6. Based on the characterization index set and evolution characteristics of the shimmy response increment, the influence of geometric characteristics on the shimmy response increment is obtained. The specific process is as follows:
[0044] Two types of continuous three-dimensional deformation representation forms are considered: ① Using simple harmonic wave signals of different wavelengths to represent continuous three-dimensional deformation, and analyzing the influence of three-dimensional deformation in different bands on the response increment. Simple harmonic wave signals include two variable parameters: wavelength and amplitude. ② Using power spectral density PSD and coherence function Coh to represent continuous three-dimensional deformation, and analyzing the influence of random three-dimensional deformation on the response increment. Based on the two types of continuous three-dimensional deformation representation forms, the three-dimensional deformation working conditions are established considering the sensitive frequency of the aircraft and the actual flatness of the airport. Based on the different established working conditions and combined with the ground motion state of the aircraft, the response increment time-frequency analysis of the swing evolution process is carried out to obtain the influence of three-dimensional deformation with different geometric characteristics on the response increment.
[0045] S7. Based on the influence of geometric characteristics on the swing response increment, a mapping model between the swing response increment and the three-dimensional deformation is established to study the enhancement effect of the runway three-dimensional deformation on the swing evolution process.
[0046] Based on the influence of three-dimensional deformation geometric characteristics on response increment, a mapping model between the two is established for practical applications: the independent variables are the three-dimensional deformation type and parameters, the dependent variable is the response increment, and the influencing factors are the aircraft ground motion state (take-off and landing state, aircraft model, and taxiing speed).
[0047] Considering that the independent variable and the dependent variable are in a high-dimensional nonlinear correlation, in order to improve the accuracy of the mapping model, this embodiment adopts a multivariate kernel ridge regression model, such as Figure 5 As shown in the figure: kernel function is used for nonlinear mapping, and ridge regression regularization is used to solve the complexity caused by the high dimensionality of the model. Given a training dataset of three-dimensional deformation type, parameters, and response increment in is the input 3D deformation feature vector, used as the training data point, is the corresponding response increment target value. The goal of kernel ridge regression is to learn the function f(x):
[0048]
[0049] In formula (4), f(x) represents the decision boundary in the support vector machine; α i is the training data point x i The weight α i , these weights are determined during the model training process; k(x i ,x j ) This is a kernel function that is used to calculate the input x and the training data point x i The kernel function enables the model to find nonlinear relationships between data in a higher-dimensional space.
[0050] In formula (5), is the loss function, which consists of two parts: the first part is the mean square error, which is used to measure the model prediction and the true value y i The second part is the regularization term λ||α|| 2 , used to control the complexity of the model weights, α is the regularization parameter, used to control the complexity of the model and prevent overfitting, the regularization term λ||α|| 2 This ensures that the weights do not become too large, thereby improving the generalization ability of the model.
[0051] Through formulas (4) and (5), the prediction on the training data is made as close to the target value as possible while maintaining the smoothness and generalization ability of the model. Then, a mapping model between the response increment and the three-dimensional deformation under different ground motion states is established, and the three-dimensional deformation of the runway is evaluated based on the mapping model.
Claims
1. A runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect, characterized in that: include: Establish a detailed simulation model of the entire aircraft; Establish a three-dimensional deformable solid model of the runway; A runway-aircraft taxiing dynamics simulation model for shimmy analysis is established based on the aircraft full-machine refined simulation model and the runway three-dimensional deformation solid model; According to the runway-aircraft taxiing dynamic simulation model, an enhancement mechanism of the shimmy evolution process is obtained; According to the enhancement mechanism of the shimmy evolution process, a characterization model and an index set for describing the shimmy response increment are established; According to the characterization index set of the shimmy response increment, the influence of geometric characteristics on the shimmy response increment is obtained; According to the influence of geometric characteristics on the shimmy response increment, a mapping model between the shimmy response increment and the three-dimensional deformation is established. The three-dimensional deformation of the runway is evaluated based on the mapping model.
2. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 1, characterized in that: The enhancement mechanism for analyzing the shimmy evolution process includes: setting two scenarios with and without continuous excitation of three-dimensional deformation, carrying out three-dimensional deformation-aircraft full-process dynamic simulation, and obtaining the time-history response of the strut torsion angle and the lateral bending angle; then, using the periodogram method to obtain the power spectral density diagram of the time domain signal, using Poincaré mapping to obtain the Poincaré section and phase diagram, analyzing the enhancement effect of the landing gear vibration response from multiple angles, and obtaining the divergence and convergence mechanism of the shimmy evolution process under the forced action of the runway three-dimensional deformation.
3. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 1, characterized in that: The method of establishing a characterization model and an index set for describing the incremental shimmy response comprises: referring to the characterization principles of the time domain, frequency domain, and energy of non-stationary signals, and establishing a characterization model and an index set for describing the incremental shimmy response according to the incremental characteristics of the support torsion angle, lateral bending angle, and wheel shimmy angle responses during the shimmy evolution process under different three-dimensional deformation excitations, through collinearity and significance analysis.
4. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 1, characterized in that: The influence of the geometric characteristics on the swing response increment includes: establishing two types of continuous three-dimensional deformation representation forms: ① using simple harmonic wave signals of different wavelengths to represent the continuous three-dimensional deformation, and analyzing the influence of three-dimensional deformation in different bands on the response increment, and the simple harmonic wave signal includes two variable parameters, wavelength and amplitude; using power spectral density PSD and coherence function Coh to represent the continuous three-dimensional deformation, and analyzing the influence of random three-dimensional deformation on the response increment; through the two types of continuous three-dimensional deformation representation forms, as well as the sensitive frequency of the aircraft and the actual flatness of the airport, the runway three-dimensional deformation working condition is established, based on the established different working conditions, and combined with the ground motion state of the aircraft, the response increment time-frequency analysis of the swing evolution process is performed to obtain the influence of three-dimensional deformation with different characteristics on the response increment.
5. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 1, characterized in that: The mapping model between the shimmy response increment and the three-dimensional deformation is established by: based on the influence of the three-dimensional deformation geometric characteristics on the response increment, a mapping model between the two is established for practical applications: the independent variables are the three-dimensional deformation type and parameters, the dependent variable is the response increment, and the influencing factor is the ground motion state of the aircraft. A multivariate kernel ridge regression model is adopted: the kernel function is used for nonlinear mapping, and the complexity caused by the high dimension of the model is solved by regularization of ridge regression. Given a training data set of three-dimensional deformation type, parameters, and response increment, in is the input 3D deformation feature vector, is the corresponding response increment target value. The goal of kernel ridge regression is to learn the function f(x). Through formulas (4) and (5), a mapping model between the response increment and the three-dimensional deformation under different ground motion states is established, and then the three-dimensional deformation of the runway is evaluated: In formula (4), f(x) represents the decision boundary in the support vector machine; α i is the data point x i The weight α i ; k(x i ,x j )This is a kernel function; In formula (5), is the loss function, which consists of two parts: the first part is the mean square error, which is used to measure the model prediction and the true value y i The second part is the regularization term λ||α|| 2 .
6. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 1, characterized in that: Import the flexible aircraft fuselage model and the flexible landing gear model into the ADMA / Aircraft virtual prototype software platform to assemble the whole aircraft assembly, obtain a detailed simulation model of the whole aircraft, and realize the ground motion behavior simulation of the full-scale aircraft; The establishment of the flexible landing gear model includes: using the finite element method to establish the strut outer cylinder and piston rod model in HyperMesh; using the linear elastic constitutive relationship and dividing the HEX grid to improve the calculation efficiency; the landing gear of the simulation model adopts a single-cavity variable oil hole buffer; for the vibration analysis, a complex tire force model is established based on the point contact theory, and the tire lateral force Flat is a function of the tire vertical force Fn and the tire side slip angle α; The establishment of the flexible aircraft fuselage model includes: based on the flexible front landing gear and main landing gear models, using a finite element model to replace the rigid fuselage, establishing a flexible fuselage model, and assembling a full-aircraft shimmy dynamics model.
7. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to claim 6, characterized in that: The establishment of the three-dimensional deformation entity model of the runway includes: for non-stationary signals of geometric mutation and continuous deformation, the mutation height Δh, the lateral and longitudinal wavelengths wt, and the amplitude A are used to characterize; for stationary signals, the random spectrum power spectrum density PSD and the coherence coefficient Coh are used to characterize, and a mathematical characterization model is abstracted, and then the three-dimensional deformation point cloud data of the runway is generated according to the mathematical characterization model, and the data is input into the MATLAB / Simulink module to generate the three-dimensional coordinates of each node, wherein the z coordinate is the three-dimensional elevation of the road surface, and three adjacent nodes are connected into triangles, and the adjacent triangles have common edges and are connected to each other to form the three-dimensional deformation entity model of the runway.
8. The runway three-dimensional deformation evaluation method based on landing gear shimmy enhancement effect according to any one of claims 1 to 7, characterized in that: In order to ensure the accuracy of the runway-aircraft taxiing dynamic simulation model in the analysis of the three-dimensional deformation swing vibration force, the simulation model is verified as follows: the runway-aircraft taxiing dynamic simulation model is built based on the ADAMS software platform, and the initial calculation step length, aircraft taxiing speed and other parameters are set. The dynamic response indicators of the swing vibration frequency and amplitude are output, and the objective function F(θ) is constructed in combination with the measured airborne perception data as shown in equations (1) to (3): F(θ)=β1VF o (θ)+β2VA o (i) (1) Where: θ represents the simulation model parameter vector; VF o Indicates frequency difference index; VA o Represents the amplitude difference index; f sr Indicates the simulation calculation frequency; f mr Indicates the measured frequency; a sr Represents the simulation calculation amplitude; a mr Indicates the measured amplitude.
