41 results about "Flight test data" patented technology

The present application belongs to the technical field of hypersonic wind tunnel test, and discloses a test checking method of hypersonic frictional resistance standard value. The test checking method of hypersonic frictional resistance standard value is carried out in a hypersonic quiet wind tunnel, and the model is stably kept in a laminar flow state by means of a quiet flow field provided by the hypersonic quiet wind tunnel, so that the interference of flow field noise on the hypersonic frictional resistance is eliminated, real near space flight test data is obtained, and the obtained hypersonic frictional resistance standard value can be used for related test calculation or theoretical analysis, thereby providing technical support for the research on the frictional resistance of a spacecraft.

The application discloses a helicopter flexible beam actual load substitution analysis method, comprising the following steps: considering that the flexible beam root actual measured load data is effective in the initial flight test, using the effective time period corresponding to the effective flight test data of the flexible beam root actual measured load, performing spectrum analysis on the flexible beam root profile actual measured load and the airfoil section profile actual measured load on the same piece of tail rotor blade at preset interval time in the effective time period of the effective flight test data, so as to obtain the amplitude of the flexible beam root profile actual measured load and the airfoil section profile actual measured load corresponding to the tail rotor rotating speed frequency respectively; combining the amplitude of each time point in the effective time period to form two groups of data columns; a scatter plot of the two groups of data columns can be drawn, a plurality of correlation formulas of the two groups of data columns are obtained through linear fitting, the correlation formula with the optimal correlation between the airfoil section profile actual measured load and the flexible beam root profile actual measured load on the same piece of tail rotor blade is selected as the basis for calculating the flexible beam root profile actual measured load, and the flexible beam root profile actual measured load in the whole flight process can be obtained based on the airfoil section profile actual measured load of the tail rotor blade.

The invention provides a flight parameter identification method based on an intelligent resolving model of an atmospheric sensing system, which comprises the following steps of: constructing the intelligent resolving model of the atmospheric sensing system by combining flight test data and ground data, and identifying flight parameters according to the intelligent resolving model of the atmospheric sensing system. According to the method, the difference between flight and ground prediction is considered, and especially for an aircraft with multiple flight test data, more flight test data are added, so that the resolving precision of an atmosphere sensing system can be effectively improved.

This invention provides an online monitoring method and system for airborne flight runaway precursors of civil aircraft based on flight test data. The method includes: constructing a set of input flight parameters for identifying airborne flight runaway precursors; acquiring daily operational data and flight test data of the target aircraft model; and extracting physical feature information of airborne flight runaway by referencing the aerodynamic mechanism model and extreme flight envelope boundary of the target aircraft model. Based on the domain adaptation concept, an offline precursor recognition model is constructed that integrates physical feature information embedding, meta-learning, and multi-instance learning. Based on knowledge distillation technology, the recognition capability of the offline precursor recognition model is transferred to a lightweight network constructed from gated recurrent units to generate an online precursor monitoring model, enabling real-time precursor probability calculation and early warning. Finally, an intelligent agent model based on a dual-delay deep deterministic policy gradient algorithm is constructed to verify the effectiveness of the online precursor warning. This invention overcomes the cross-domain data gap and meets the requirements of online lightweight computation.

The invention relates to a flight test telemetering data processing method, which belongs to the technical field of flight test data processing, and comprises the following steps of: firstly, loading and preprocessing a flight test telemetering data file, drawing a curve of a measurement value of each measurement channel along with time change, and outputting a resolving data file; specifying an interpolation method and a time sequence, and outputting an interpolation table; the time sequence of the interpolation table is converted into a monotone sequence, preprocessing is carried out, timestamp verification is carried out, outlier data are marked, 3 sigma verification is carried out, abnormal data are eliminated, interpolation is carried out on the eliminated abnormal point data according to previous and later data, and recalculation is carried out on the obtained interpolation table according to existing telemetry data; a data cleaning rule is given, and a result is output after data cleaning. According to the method, accurate synchronization of the time sequence of the heterogeneous system can be realized, multi-mode abnormal data cleaning is supported, and the defects of high-frequency information loss, poor abnormal value detection adaptability and low processing efficiency during multi-source telemetry data time sequence alignment in the prior art are overcome.

The application relates to an airplane static pressure source error correction method based on a Bayesian regularization neural network, comprising the following steps: constructing a mixed training data set, wherein the mixed training data set comprises flight test data, CFD simulation data and physical boundary data; pre-processing and dividing the mixed training data set to form a training set and a test set; constructing a Bayesian regularization neural network model and training; and using the trained Bayesian regularization neural network model to realize static pressure source error correction. The application uses CFD numerical simulation technology to economically and efficiently generate dense samples covering the whole design envelope, so as to make up for the insufficiency of flight test data in the coverage range and data density of the envelope; meanwhile, a Bayesian regularization training algorithm is introduced, the model complexity and data fitting degree are automatically balanced through Bayesian inference, overfitting is effectively inhibited, and the model generalization ability is improved.

A quantitative flight test data gathering system is configured to, during flight test operation of an aircraft, obtain, from one or more flight data sensors, from a flight control system, or both, one or more feedback signals indicative of a state of the aircraft. The quantitative flight test data gathering system is configured to generate, based at least in part on the feedback signals, one or more signals associated with a flight test maneuver. The quantitative flight test data gathering system is configured to provide the one or more signals as input to the flight control system to cause one or more control laws of the flight control system to generate control signals to cause the aircraft to operate according to the flight test maneuver.

The invention relates to the technical field of flight simulation, in particular to a flight simulation equipment control quality goodness of fit calculation method based on incomplete test flight data. Comprises: determining a test item; acquiring real flight data of the test item and simulation data of flight simulation equipment; preprocessing the real flight data and the simulation data; complementing the missing part of the real flight data; and comparing the real flight data with the simulation data, and calculating the control quality goodness of fit. According to the invention, automatic testing of flight test data and simulated flight data can be realized, and the defect of inaccuracy caused by subjective artificial evaluation is overcome.

This application provides a method, apparatus, device, storage medium, and program product for measuring fuselage parameters, belonging to the field of aviation technology. The method includes: performing a flight test to obtain target flight test data for a target area of ​​the aircraft, the flight test data including the aircraft's vertical acceleration, fuselage angular acceleration about the pitch axis, longitudinal strain variation, and circumferential strain variation; performing a ground test to obtain target ground test data for the target area, the ground test data including the coefficients of pressure circumferential stress versus pressure, pressure longitudinal stress versus pressure, and bending longitudinal stress versus bending moment; and determining local fuselage physical parameters for the target area based on the target flight test data and the target ground test data, the local fuselage physical parameters including local torque and moment of inertia. This application can reduce the cost of obtaining local fuselage physical parameters.

The invention provides a method for automatically identifying and testing objective subjects of a D-level flight simulator, which comprises the following steps of: processing and intercepting flight test data: respectively carrying out flight test data processing and intercepting work according to flight performance and flight characteristic subjects specified by a simulator objective test standard; making and generating a test configuration file, extracting test flight condition information from the intercepted flight test data, and making a configuration file; automatic identification testing: loading the configuration file and the intercepted flight test data into an existing flight simulation program of the flight simulator, respectively executing automatic identification testing according to subjects specified by a standard, and recording a test result; and automatically evaluating a test result, drawing an intercepted flight test data curve and a tolerance specified by a standard, drawing a test result curve, identifying a tolerance conforming part and an out-of-tolerance part, and giving conformity evaluation. The method ensures the legitimacy and validity of the test process and result, and meets the objective test requirements of the D-level analog machine specified by the standard.

This invention discloses a method for identifying the cruise drag coefficient of an aircraft based on flight test data. First, a longitudinal motion state space model of the aircraft containing trim information is established, and the trim information is introduced as a parameter to be identified, enabling direct input of the original full flight test data. Then, the model is decoupled into four independent single-output subsystems, and the optimal estimates of the aerodynamic derivative and trim information are obtained in one step through global optimization using the least squares method. Finally, the true trim state point is determined based on the trim information, and the time history of the cruise drag coefficient is calculated using a linear incremental drag coefficient model. This invention has strong anti-interference capabilities, does not rely on initial trim values, reduces the standard deviation of the identified drag coefficient by 14.6% compared to traditional methods, and achieves a model prediction accuracy R² of over 0.99. It can directly utilize polar curve flight test QAR data without additional flight test costs and is suitable for aerodynamic parameter identification in all flight phases of conventional layout civil aircraft.

The invention relates to an airborne flight correction evaluation method for a satellite-borne early warning radar, and the method comprises the steps: 1, carrying out the preprocessing of flight correction test data, obtaining the detection trace point number M, the radar operating distance RE, m, the signal-clutter-noise ratio margin SCNRE, m, the actual flight trace point number N of a target aircraft, and the RCS value sigma E, m, and m = 1, 2,... M in a flight correction test, and obtaining the detection trace point number M, the radar operating distance RE, m, the signal-clutter-noise ratio margin SCNRE, m, the actual flight trace point number N of the target aircraft, and the RCS value sigma E, m; 2, calculating the on-orbit detection distance RG, m of the satellite-borne early warning radar according to the radar operating distance RE, m; and step 3, carrying out balance conversion between the detection distance and the discovery probability to obtain an on-orbit detection distance R under the condition of a specific discovery probability p. According to the airborne flight correction evaluation method for the satellite-borne early warning radar, the problems of conversion between the clutter-noise ratio margin and the operating distance and between the discovery probability and the radar operating distance are comprehensively considered, flight correction test data information is fully mined, and the accuracy and credibility of calculation evaluation are improved.

The invention discloses an aircraft control system optimization method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring actually measured test flight data acquired by the multi-rotor aircraft in a test flight process; performing time-domain state analysis on the time-measured flight test data, and determining a time-domain analysis result; performing low-order fitting processing on the time actual measurement flight test data, determining a single-channel identification model corresponding to the time multi-rotor aircraft, performing frequency domain state analysis on the time single-channel identification model, and determining a frequency domain analysis result; when at least one of the time-domain analysis result and the time-frequency domain analysis result does not pass verification, obtaining a parameter configuration file carrying target control law parameters; and deploying the hour parameter configuration file to the aircraft control system on line, so that the hour aircraft control system continues to carry out test flight based on the target control law parameters. According to the method, the control law parameters can be iteratively updated online, and the iteration efficiency and accuracy of the control law parameters are greatly improved.

The invention belongs to the technical field of aircraft test data processing, and discloses an aircraft-oriented test data comprehensive analysis method and system, and the method comprises the steps: collecting aircraft test data, including engineering simulator data, test flight data and laboratory test data, and carrying out the preprocessing of the aircraft test data, the method comprises the following steps: establishing a data association model according to a sensor signal state of aircraft test data, airborne system and equipment state information and alarm logic, associating the aircraft test data, extracting alarm information, abnormal modes and fault features of the associated aircraft test data, generating a data chart related to fault alarm, and if the alarm information does not exist, executing the fault alarm; if yes, generating an abnormal data chart and generating an analysis report. According to the invention, multi-type data can be efficiently processed, faults can be rapidly positioned, data can be visually displayed, and aircraft maintenance is supported.

The invention relates to the technical field of data analysis, in particular to a test flight data analysis method and system based on a cloud platform, and the method comprises the following steps: obtaining data, sorting logs according to sorties and timestamps, judging an undercarriage, speed and normal overload based on a time sequence event and a cumulative duration rule, generating a key value pair, and calculating the test flight data; the grouping event key value pair is input into the finite state automaton to output profile violation, the grouping duration key value pair calculates the total overrun duration to generate an accumulated overrun report, and the accumulated overrun report is combined to form a comprehensive report. According to the method, a time sequence log is constructed by collecting test flight sorties and timestamps, automatic data regulation is achieved, the undercarriage state, the flight speed and accumulated normal overload are judged by means of a time sequence event association rule, manual scripts are replaced, an event structure is input into a finite state automaton, violation logic state transition is matched, and the reliability of the undercarriage is improved. According to the method, the violation of the complex section is automatically identified, the total accumulated overrun duration is calculated, and the comprehensive violation verification report is combined and output, so that the efficiency and accuracy of data analysis are remarkably improved.

The invention relates to the technical field of aviation test flight, in particular to a TD-DMDc and OE fusion-based flutter test flight data modal automatic identification method, which comprises the following steps of: respectively extracting modal parameters from flutter test flight data by using an OE model and a TD-DMDc method, preliminarily screening, and obtaining candidate modal pairs; according to the candidate modal pair, single-modal reconstruction is carried out through an OE model and a TD-DMDc method, and corresponding single-modal reconstruction signals are obtained; according to the single-mode reconstruction signals respectively corresponding to the OE model and the TD-DMDc method, obtaining a comprehensive similarity measurement value by calculating the correlation between the single-mode reconstruction signals; performing envelope analysis on the flutter flight test data to obtain a real modal number; and obtaining a real mode according to the comprehensive similarity measurement value and the number of the real modes. According to the method, the parametric index and the behavioral index are fused, so that a modal verification system which is more robust and closer to physical reality can be constructed.

The application discloses a kind of helicopter airborne rotor blade tip displacement measurement methods based on point light source detection, belong to image data processing field.The method will point light source be fixed in helicopter rotor blade tip area to realize all-weather blade tip mark, and industrial camera is fixed at the bottom of helicopter with the angle of inclination to collect rotor blade tip area image, in second, a kind of small target detection deep learning model is designed in combination with Gauss receptive field and weighted feature fusion, can realize the accurate identification and positioning of point light source in larger field of view.The method mainly includes the following steps:1) visual calibration of industrial camera;2) rotor blade tip area image acquisition;3) point light source detection based on deep learning;4) oblique vision error compensation, and calculate blade tip displacement amount.The application uses deep learning network model and computer vision to realize helicopter airborne rotor blade tip displacement measurement, with the advantages of high precision, real-time, all-weather, can provide flight test data support for new helicopter design.

The invention discloses a high-altitude rocket rarefied flow field reconstruction method and device based on flight test data. The method comprises the following steps: constructing a high-altitude rocket rarefied flow field reconstruction model and a Boltzman-BGK equation non-equilibrium term prediction model based on a Shakhov equilibrium state distribution function; the unbalanced term prediction model comprises a first sub-network and a second sub-network which are arranged in parallel and respectively output an unbalanced term of a density distribution function and an unbalanced term of a temperature distribution function; a loss function used in the training process comprises a first loss function constructed based on a residual term of a Boltzmann-BGK equation, and a second loss function constructed based on a residual term of the Boltzmann-BGK equation. The second loss function is constructed on the basis of a residual term of physical constraint represented by an unbalanced term; the third loss function is constructed on the basis of flight test data and a predicted value of flow field data output by the high-altitude rocket rarefied flow field reconstruction model; and inputting a to-be-predicted flow field coordinate on the high-altitude rocket into the trained high-altitude rocket rarefied flow field reconstruction model to obtain a predicted value of the flow field data. According to the invention, the flow field reconstruction precision can be improved.

The invention provides a carrier rocket safety margin design method and system based on probability calculation and a processor. According to the invention, through probability statistical calculation, a forward numerical design method for directly solving the safety margin based on the known depletion shutdown / injection probability is realized. The method does not depend on accumulation of historical design experience, the safety margin can be directly obtained through modeling calculation, and therefore the method is especially suitable for research and development of brand new rockets lacking reference models. According to the method, by adopting the forward design, the selection of the traditional'safety margin cluster 'and a large amount of random targeting iteration are fundamentally avoided, so that the design period is remarkably shortened, and the design efficiency is improved. The method provided by the invention has been successfully applied to the safety margin design of various models of carrier rockets, and multiple simulation results and flight test data prove the effectiveness of the method.

The invention belongs to the technical field of aircraft test flight data processing, and discloses a test flight data analysis method based on a node graph, which comprises the following steps of: analyzing a processing flow of test flight data, creating a node graph of test flight data analysis, defining a name and a function of each node, and creating input and output of each node in the node graph. And compiling a test flight data processing task program of each node, constructing an execution sequence of the test flight data processing task programs based on the node graph, and completing tasks on all the nodes of the test flight data according to the execution sequence. According to the method, by creating the data processing workflow, the data processing process is easy to master, the data processing flow is standardized, in addition, in the creation process, various data processing nodes can be developed, and the nodes can be shared to others for use and can also be reserved as technology accumulation; in addition, a plurality of data processing workflows can be created in software, and the workflows are processed at the same time through a multi-thread technology, so that the data processing efficiency is improved.

This invention belongs to the field of flight test data processing technology and discloses a method and system for quality assessment of multi-source flight test data based on physical-temporal constraints. The method includes the following steps: data preprocessing and benchmark calibration; physical-temporal joint verification; multi-source spatiotemporal alignment; scenario adaptive evaluation; and comprehensive quality scoring and feedback optimization. This invention solves the problems of insufficient correlation and inconsistent spatiotemporal benchmarks in existing flight test multi-source data, achieving full-dimensional and high-precision quality assessment of multi-source flight test data.

The invention belongs to the technical field of unmanned aerial vehicle design, and particularly relates to a longitudinal aerodynamic parameter identification calculation method under a closed-loop state and a small attack angle range, and the method comprises the steps: simulating the flight state change process of an unmanned aerial vehicle under the condition of the closed-loop state and the small attack angle range based on a flight simulation model; dynamic characteristics of the aircraft under different longitudinal pneumatic conditions are obtained; performing parameter identification calculation on the dynamic characteristics of the aircraft under different longitudinal pneumatic conditions to obtain flight simulation data; iterative optimization is performed on the longitudinal aerodynamic model parameter combination set through a genetic algorithm, and an optimal longitudinal aerodynamic model parameter combination is obtained by combining the sum of squares of errors between flight simulation data and actual flight test data; and simulation verification and error analysis are carried out on the optimal longitudinal aerodynamic model parameters, and verification is completed when a preset precision requirement is met. The influence of the control law on the flight test in the flight process is considered, and the identification of the longitudinal aerodynamic coefficient parameters in the corresponding closed-loop state is more accurate.

The present application belongs to the technical field of helicopter design, and particularly relates to a method for determining the flight performance of a helicopter through flight test data. The method comprises: obtaining the hovering required power according to the hovering flight test data; determining the level flight required power according to the variable-height level flight test data, wherein the helicopter equivalent weight is kept unchanged by adjusting the flight height during the variable-height level flight; determining the climb efficiency and the climb required power according to the sawtooth climb flight test data; determining the engine installation loss according to the hovering, level flight and climb flight test data, so as to determine the available power of the helicopter; and determining the flight performance according to the required power and the available power of the helicopter in the hovering, level flight and climb states.

The invention relates to an aircraft, a method of collecting flight test data, and a computer readable medium. A quantitative flight test data collection system is configured to acquire, during a flight test operation of an aircraft, one or more feedback signals indicative of a state of the aircraft from one or more flight data sensors, from a flight control system, or from both. The quantitative flight test data collection system is configured to generate, based at least in part on the feedback signals, one or more signals associated with a flight test maneuver. The quantitative flight test data collection system is configured to provide the one or more signals as input to the flight control system to cause one or more control laws of the flight control system to generate control signals to cause the aircraft to operate in accordance with the flight test maneuver.

The invention provides a test flight action automatic identification method and device, and the method comprises the steps: obtaining test flight data which is a data sampling point set with time information; acquiring a to-be-screened action set from the test flight data by adopting a preset sliding window; the duration t of the preset identification sliding window is K times of the single pulse manipulation time t0; the duration of the action to be screened is t; the manipulated variable amplitude absolute value of the to-be-screened action is greater than a first threshold value; the peak-peak value time of the action to be screened is smaller than a second threshold value; the 0-axis crossing frequency of the to-be-screened action is not greater than a third threshold value; and obtaining a cross correlation coefficient of each to-be-screened action in the to-be-screened action set and the standard multiple pulse action, and identifying the multiple pulse excitation action in the to-be-screened action set according to the condition that the maximum value of a cross correlation coefficient vector is not less than a fourth threshold value. Multiple pulse excitation test flight actions can be automatically identified, and the automation degree, execution efficiency and readability of flight test data processing are improved.

The invention discloses a real-time calculation method for a synthetic parameter algorithm chain based on real-time data, which belongs to the technical field of data processing and comprises the following steps: S1, a data integration stage, S2, a data storage stage and S3, a synthetic parameter calculation stage. According to the method, the aviation test flight data monitoring efficiency is remarkably improved, flight abnormity is rapidly captured through a millisecond-level real-time computing architecture, the response speed is increased by ten times compared with a traditional system, a parameter formula is dynamically edited and synthesized in test flight and takes effect immediately, an open type analysis engine is combined to be compatible with multi-language algorithm extension, and the system is high in reliability. A two-stage storage mechanism guarantees zero data loss during faults, a preprocessing layer automatically immunizes abnormal data interference, all step results of an algorithm chain are visually displayed in a whole process, accurate positioning of abnormal links is assisted, a historical data rapid playback function supports scene reproduction and parameter adjustment and optimization, and meanwhile the system coupling degree is reduced through a data-algorithm decoupling design. The newly added sensor does not need to reconstruct a calculation link.

The invention discloses a carrier rocket flight test swing nozzle physical swing angle estimation method and a carrier rocket flight test swing nozzle physical swing angle estimation system. The carrier rocket flight test swing nozzle physical swing angle estimation method comprises the following steps: acquiring servo mechanism geometric parameters; acquiring flight test data; determining the absolute length of the two-channel actuator in the flight process; estimating the sinkage of the swing center when the engine works according to the obtained flight test data; according to the estimated sinkage of the swing center when the engine works and the absolute length of the two-channel actuator, the actual physical swing angle of the swing spray pipe is determined; determining an angular displacement length estimation result according to the actual physical swing angle of the swing nozzle and the estimated sinkage of the swing center when the engine works; judging whether the angular displacement length estimation result meets the requirement or not; if the requirement is met, outputting the actual physical swing angle of the swing spray pipe; on the contrary, the sinkage of the swing center is corrected. The accuracy of the calculation result is improved, and a reliable basis can be provided for evaluation of the actual dynamic characteristics of the swing nozzle.

The invention relates to the technical field of test flight data processing, and discloses a test flight data mining method based on a machine learning algorithm. Fusing the data; analyzing online learning and offline learning; and outputting a result. By acquiring multi-source data, calculating data reliability weight, setting a real-time data verification and retransmission mechanism, optimizing data interpolation filling in combination with a historical trend and eliminating dimensional interference by adopting Z-score standardization, the influence of an external environment on the flight performance of an aircraft is comprehensively considered, the accuracy and credibility of a test flight data mining result are improved, and the test flight data mining efficiency is improved. A scene machine learning basic model is selected according to a test flight task type, model parameters are dynamically updated in combination with an online gradient descent algorithm, a traditional offline processing mode is replaced, rapid capture of real-time test flight state changes is achieved, potential problems and abnormities are found in time, test flight safety is guaranteed, and test flight efficiency is improved.

The invention relates to a flight test pneumatic identification system and method, and belongs to the technical field of aircraft test data post-processing. A data processing layer module receives flight test data and outputs flight test parameters to a core algorithm layer module; receiving reference data to form an interpolation table; the core algorithm layer module comprises a gravity compensation module and is used for calculating a real-time gravity vector and sending the real-time gravity vector to the kinematics resolving module; the attitude control force compensation unit is used for reading steering engine deflection parameters, outputting control force and torque and sending the control force and the torque to the kinematics resolving module; the kinematics resolving module calculates and outputs aerodynamic force and torque coefficients; the multi-dimensional data interpolation module is used for receiving the flight test parameters sent by the data processing layer module, reading an interpolation table in the data processing layer module and carrying out interpolation based on the interpolation table to obtain aerodynamic force and torque coefficients; and the verification output layer module performs merging according to the time sequence of the flight test measurement data to obtain a comparison curve of the test data and the reference data.

The invention belongs to the technical field of unmanned aerial vehicle aircraft design, and particularly relates to a method for identifying and calculating parameters of a longitudinal aerodynamic model in an open-loop state and a small-angle-of-attack range, and the parameters of the longitudinal aerodynamic model of an aircraft are obtained. Firstly, longitudinal aerodynamic model parameters needing to be identified are determined, and an initial longitudinal aerodynamic model parameter combination set is formed; performing flight simulation on the aircraft under different longitudinal aerodynamic model parameter combinations by adopting a combinatorial optimization method, and taking an error sum of squares between flight simulation data and actual flight test data as a fitness function; performing iterative optimization on the longitudinal aerodynamic model parameter combination set through a genetic algorithm to obtain an optimal longitudinal aerodynamic model parameter combination in the current longitudinal aerodynamic model parameter combination set; and performing simulation verification and error analysis on the optimal longitudinal aerodynamic model parameter combination, and determining that a result meets preset precision. The method has universality and is suitable for parameter identification of longitudinal aerodynamic model parameters of different types of unmanned aerial vehicles.