A method for evaluating dynamic characteristics of a launch vehicle flight test

Abstract

This invention discloses a method for evaluating the dynamic characteristics of a launch vehicle during flight tests. This method can assess the dynamic characteristics of a launch vehicle under actual flight conditions based on flight test results. As a crucial input to the attitude control system, the evaluation results of the rocket's dynamic characteristics can be used to further refine the rocket's dynamic model, reveal the transmission paths of key components, and improve the predictive accuracy of the rocket's dynamic characteristics for subsequent launches. This method for evaluating the dynamic characteristics of launch vehicles during flight tests is a general-purpose technique that can be widely applied to various launch vehicle models.

Description

Technical Field

[0001] This invention relates to the field of launch vehicles, specifically to a method for evaluating the dynamic characteristics of a launch vehicle with bundled solid boosters during flight tests, which can evaluate the dynamic characteristics of the launch vehicle under actual flight conditions based on flight test results. Background Technology

[0002] The dynamic characteristics of the rocket body are one of the key aspects of the overall design of the launch vehicle. The bending, torsion, and longitudinal characteristics of the entire rocket are used for elastic design, which is used by the attitude control and payload design professionals. The lateral and longitudinal transmission characteristics of the entire rocket are used for star-rocket coupling analysis and POGO stability design analysis.

[0003] Existing modal analysis methods are all based on data from ground-based modal tests. However, the boundary conditions and stress states of a launch vehicle under ground test conditions differ from those under actual flight conditions, allowing for only approximate simulations or attributing minimal impact on the identification of the structure's inherent characteristics. Furthermore, existing modal analysis methods require the measurement of both excitation and response. Since a launch vehicle's flight state is a typical "output-only" system, sensors mounted on the rocket can only measure the response, making it difficult to measure the excitation. Therefore, traditional modal analysis methods cannot be used to analyze flight test results.

[0004] Since the measured response is caused by unknown or known forces acting at discrete locations on the structure (e.g., engine excitation, pulsating pressure) or distributed throughout the structure (e.g., wind excitation), there is no practical way to confirm whether all modes of the system, or the modes of interest, are excited. If the unknown force spectrum cannot excite all modes, the system response may not accurately reflect all frequencies. If any excitation acts on a node, the system response will certainly be unobservable. Furthermore, it is necessary to distinguish between the structure's own frequencies and the forced vibration frequencies caused by external excitations. Summary of the Invention

[0005] The purpose of this invention is to provide a method for evaluating the dynamic characteristics of a launch vehicle during flight tests. This method uses measurement data of the launch vehicle when it is in actual working condition to extract the dynamic characteristics of the rocket body.

[0006] To achieve the above-mentioned objectives, this invention provides a method for evaluating the dynamic characteristics of a launch vehicle during flight tests. This method, based on flight test results, can evaluate the dynamic characteristics of a launch vehicle under actual flight conditions. Specifically, it is implemented through the following steps: The method for evaluating the dynamic characteristics of a launch vehicle during flight tests, based on time-domain telemetry data acquired during launch vehicle flight tests, includes the following steps:

[0007] Step 1, remove outliers;

[0008] Step 2: Centralize the signal data processed in Step 1;

[0009] Step 3: Resample the signal data processed in Step 2:

[0010] Step 4: Select data from the signal data processed in Step 3 and construct a model using nodes and connections;

[0011] Step 5: Select at least one node in the model built in step 4 as a reference point for modality identification;

[0012] Step 6: Calculate the cross-power spectral density between the signal data in the model in Step 4 and the reference point signal data in Step 5.

[0013] Step 7: Perform parameter estimation on the cross power spectral density after step 6, and extract the frequency, damping, and mode dynamic characteristics.

[0014] Preferably, in step 1, outlier removal: outliers are removed using a threshold method. For points exceeding the threshold, interpolation is performed, and the data of two adjacent points of the outlier are selected for interpolation.

[0015] Preferably, in step 2, centering (detrending term): Centering generally involves subtracting the DC bias or mean of the signal. Detrending is achieved using polynomial fitting or a moving average. When using polynomial fitting, the polynomial order must be carefully selected to avoid underfitting or overfitting. For low-frequency signals, centering is generally used without detrending; for slowly varying signals such as pressure signals, detrending is necessary due to the obvious trend.

[0016] Preferably, in step 3, resampling: for low-frequency signals with a high sampling rate, a lower sampling rate is selected to resample the signal. Additionally, to minimize aliasing, a low-pass filter must be introduced before resampling.

[0017] Preferably, in step 4, model construction: select appropriate measurement point signals, define the nodes and connections of the measurement points, and create a geometric model.

[0018] Preferably, in step 5, reference point selection: among the selected measurement point signals, one or more measurement point signals are selected as reference points for modal identification.

[0019] Preferably, step 6, cross-spectrum calculation: calculate the cross-spectrum between the selected measurement point signal (excluding the reference point signal) and the reference point signal.

[0020] Preferably, in step 7, parameter estimation: the modal parameter estimation method adopts the least squares complex exponential method (the system's time-domain response consists of a series of damped, exponentially decaying sine functions). Based on cross-spectrum extraction frequency fi Damping b i and mode Isokinetic properties.

[0021] The present invention provides a method for evaluating the dynamic characteristics of a launch vehicle during flight tests. This method can assess the dynamic characteristics of a launch vehicle under actual flight conditions based on flight test results. As a crucial input to the attitude control system, the evaluation results can be used to further refine the launch vehicle's dynamics model, reveal the transmission paths of key components, and improve the predictive accuracy of the launch vehicle's dynamic characteristics for subsequent launches. This method for evaluating the dynamic characteristics of launch vehicles during flight tests is a general-purpose technique that can be widely applied to various launch vehicle models. Attached Figure Description

[0022] The following embodiments and accompanying drawings illustrate a method for evaluating the dynamic characteristics of a launch vehicle during flight tests.

[0023] Figure 1 The analysis consists of the original low-frequency radial vibration signal of the first-level inter-box section and the 0-10 second segment extracted as the signal.

[0024] Figure 2 It is the signal after removing outliers in the radial direction of the first-level inter-box section from 0-10 seconds.

[0025] Figure 3 It is the signal after radial 0-10 second centralization processing of the first-level inter-box section.

[0026] Figure 4 It is the signal after radial resampling of the first-level inter-box section from 0 to 10 seconds.

[0027] Figure 5 It is a steady-state diagram based on the identification results of the radial reference point of the first-level inter-box section.

[0028] Figure 6 It is the first-order mode shape of the radial direction of the arrow body.

[0029] Figure 7 A flowchart of a method for evaluating the dynamic characteristics of a launch vehicle during flight tests is provided. Detailed Implementation

[0030] The following is a further detailed description of a method for evaluating the dynamic characteristics of a launch vehicle flight test according to the present invention, with reference to the accompanying drawings.

[0031] The present invention provides a method for evaluating the dynamic characteristics of a launch vehicle during flight tests, comprising the following steps:

[0032] Step 1, remove outliers;

[0033] Step 2: Centralize the signal data processed in Step 1;

[0034] Step 3: Resample the signal data processed in Step 2:

[0035] Step 4: Select data from the signal data processed in Step 3 and construct a model using nodes and connections;

[0036] Step 5: Select at least one node in the model built in step 4 as a reference point for modality identification;

[0037] Step 6: Calculate the cross-power spectral density between the signal data in the model in Step 4 and the reference point signal data in Step 5.

[0038] Step 7: Perform parameter estimation on the cross power spectral density after step 6, and extract the frequency, damping, and mode dynamic characteristics.

[0039] The implementation plan for each step is explained in detail below.

[0040] Step 1: Remove outliers. The specific implementation steps are as follows: Use a threshold method to remove outliers. For points exceeding the threshold, use interpolation. The interpolation is performed by selecting the data of two adjacent points of the outlier.

[0041] In this example, the low-frequency signal from the radial measurement point of the first-level inter-box section is selected as the subject of this description. Figure 1 The display shows the original radial signal of the first-level inter-box section and the signal extracted from 0-10 seconds for analysis. Figure 2 Displays the signal after removing outliers in the radial direction of the first-level inter-box section from 0 to 10 seconds.

[0042] Step 2, Centralization Processing. The specific implementation steps are as follows: For low-frequency signals such as those in the first-level inter-cell section, centralization processing involves subtracting the signal mean, without detrending. The signal obtained in Step 2 after outlier removal is subtracted from its mean, resulting in the following signal: Figure 3 As shown.

[0043] Step 3, resampling. The specific implementation steps are as follows: Perform anti-aliasing filtering on the signal processed in Step 2, filtering out frequencies above the highest frequency component of the signal of interest. The cutoff frequency is selected as 25Hz. For low-frequency signals with a sampling rate of 320Hz, select a sampling rate of 50Hz (generally set to twice the cutoff frequency) to resample the signal. Figure 4 Displays the signal after radial resampling for 0-10 seconds in the first-level inter-compartment section.

[0044] Step 4, Model Construction. The specific implementation steps are as follows: Select the following measurement points: spacecraft docking surface, second stage tank section, second stage kerosene tank, first and second stage interstage section, first stage oxygen tank, first stage tank section, first stage kerosene tank, first stage engine swing point, first stage engine level seat, booster nose cone, booster tail section, etc. Define the nodes and connections of the measurement points, and create a geometric model.

[0045] Step 5, Reference Point Selection. The specific implementation steps are as follows: Among the measurement points selected in Step 4, select a measurement point at a certain location as the reference point for modal identification. In this example, the measurement point of the first-level box section is selected as the reference point.

[0046] Step 6, Cross-spectrum Calculation. The specific implementation steps are as follows: Calculate the cross-spectrum of the signals from the measurement points at the launch-satellite docking surface, the second-stage inter-box section, the second-stage kerosene tank, the first-stage / second-stage inter-stage section, the first-stage oxygen tank, the first-stage kerosene tank, the first-stage engine swing point, the first-stage engine level seat, the booster nose cone, and the booster tail section with the signals from the measurement points in the first-stage inter-box section. The cross-spectrum is obtained by multiplying the spectrum of the measurement point signal by the conjugate of the spectrum of the reference point signal.

[0047] Step 7, Parameter Estimation. The specific implementation steps are as follows: Perform an inverse Fourier transform on the cross-spectrum obtained in Step 6 to obtain the corresponding time-domain data h. Use the least squares complex exponential method for modal parameter estimation. This method analyzes time-domain data and assumes that the system's time-domain response consists of a series of damped, exponentially decaying sine functions.

[0048] The implementation of the least squares complex exponential method is as follows:

[0049] Describe a damped sinusoidal sampled signal:

[0050]

[0051] According to Euler's formula, the complex exponential form of a signal is:

[0052]

[0053] The above formula can be expressed in the following form:

[0054] Where a i x represents the amplitude of the system. i This represents the poles of the system.

[0055] The discrete time-domain signal described above can be represented as:

[0056]

[0057] Multiply the i-th equation in the above formula by c i By combining and adding the results, we can obtain:

[0058] h(0)c0+h(1)c1+…+h(N)=0

[0059] Among them, c i Satisfying the characteristic polynomial

[0060] Similarly, we can conclude that:

[0061]

[0062] Solving the above equation using the least squares method yields the constant coefficient c. i After obtaining the constant coefficients, solve for the roots x of the characteristic polynomial. i x i These are the system's poles. Once the poles are known, the corresponding frequency and damping ratio can be calculated, along with the amplitude and phase. The sign of the amplitude can then be determined based on the phase, and the mode shape can be identified. In this example, the steady-state diagram and first-order bending mode shape based on the identification results from the radial low-frequency measurement points in the first-stage inter-box section are as follows: Figure 5 and Figure 6 As shown.

[0063] The dynamic characteristic evaluation method for launch vehicle flight tests of this invention has been successfully applied to the analysis of flight test results of a new generation of launch vehicles. The dynamic characteristic evaluation results show good agreement with ground test results and simulation analysis results in the first two stages. The method of this invention can also be applied to the development of other launch vehicle models.

[0064] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for evaluating dynamic characteristics of a launch vehicle flight test, characterized by, The aforementioned method for evaluating the dynamic characteristics of launch vehicle flight tests, based on time-domain telemetry data acquired during launch vehicle flight tests, includes the following steps: Step 1, remove outliers; Step 2: Centralize the signal data processed in Step 1; Step 3: Resample the signal data processed in Step 2: Step 4: Select data from the signal data processed in Step 3 and construct a model using nodes and connections; Step 5: Select at least one node in the model built in step 4 as a reference point for modality identification; Step 6: Calculate the cross-power spectral density between the signal data in the model in Step 4 and the reference point signal data in Step 5. Step 7: Perform parameter estimation on the cross power spectral density after step 6, and extract the frequency, damping, and mode dynamic characteristics; In step 1, a threshold method is used to remove outliers. For points that exceed the threshold, interpolation is used, and the data of two adjacent points of the outlier are selected for interpolation. The centralization process in step 2 includes: for the signal data processed in step 1, centralization process subtracts the DC bias or mean of the signal; The centering process in step 2 also includes a detrending term. The detrending term is processed using polynomial fitting or moving average. When using polynomial fitting, the polynomial order needs to be carefully selected to avoid underfitting or overfitting. For low-frequency signals, the processing method is centering, without detrending term processing. For slowly varying pressure signals, since the trend is obvious, detrending term processing is required. In step 3, for the signal data processed in step 2, for low-frequency signals with a higher sampling rate, a lower sampling rate is selected to resample the signal; In step 3, a low-pass filter is introduced before resampling to minimize aliasing; Step 4, model building, includes: selecting appropriate measurement point signals, defining the nodes and connections of the measurement points, and creating a geometric model; Step 5, the reference point selection, includes selecting one or more measurement point signals from the selected measurement point signals as reference points for modal identification; Step 6 involves calculating the cross spectrum between the selected measurement point signal and the reference point signal. Step 7 parameter estimation includes: the modal parameter estimation method adopts the least squares complex exponential method, and extracts the frequency, damping and mode dynamic characteristics based on the cross spectrum.

Images