CN115422835A - Flight landing gear actuator fatigue estimation method and system - Google Patents
Flight landing gear actuator fatigue estimation method and system Download PDFInfo
- Publication number
- CN115422835A CN115422835A CN202211049584.2A CN202211049584A CN115422835A CN 115422835 A CN115422835 A CN 115422835A CN 202211049584 A CN202211049584 A CN 202211049584A CN 115422835 A CN115422835 A CN 115422835A
- Authority
- CN
- China
- Prior art keywords
- service
- actuator
- model
- prediction
- fatigue
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention provides a fatigue estimation method and a system for an actuator of a flight undercarriage, which comprises the following steps: acquiring a service history sample set of a flight undercarriage actuator; preprocessing a service history sample set, and marking a service state; respectively adopting GRNN, RBF and DNN for training to obtain a plurality of actuator fatigue estimation models, and outputting the models as actuator fault signs; counting the prediction precision value of each model in a preset service time period; carrying out weighted summation on the prediction precision value of each model in a preset service time period, and determining weighted prediction precision; obtaining a model with the highest weighted prediction precision as a final fatigue estimation prediction model; and taking the characteristic vector of the aircraft as an input, inputting the characteristic vector into the fatigue estimation prediction model, and outputting a predicted actuator fault sign. The method can be used for fault early warning and warning of the actuator of the aircraft landing gear in a service state, and the defects of manual judgment and subjective judgment are reduced.
Description
Technical Field
The invention relates to the technical field of flight simulators, in particular to simulation of an actuator of a flight undercarriage, and particularly relates to a fatigue estimation method and system for the actuator of the flight undercarriage.
Background
The actuator is a main executive component of a hydraulic transmission system, is applied to a plurality of important positions such as elevators, rudders, hatches, undercarriages and the like on the aircraft, plays an important role in a steering engine control system, a steering system and a power control system, and has direct relation between the performance and the safety of the actuator and the aircraft. The working performance of the actuator and whether the actuator is in service safely directly influence the performance and safe flight of the aircraft.
The landing gear of the aircraft is the key for safe take-off and landing, and in the design of adopting a front three-point support, a rear three-point support or multiple rows and groups, the landing gear bears the main weight of the aircraft, and the flight safety is directly influenced by high-speed running and smooth take-up and pay-off during take-off and landing. In the fields of civil aircraft and military aircraft, crash/forced landing accidents caused by landing gear faults exist for many times, when the aircraft lands at one brake, tens of tons or hundreds of tons of pressure borne by the wheels are directly applied to the landing gear, and the main shock absorber and the retractable actuator are relied on, if the hydraulic support is in service for a long time, invisible risks or damage exist, finally, the contact between the aircraft body and the wings is caused, and flight safety accidents are caused.
In a design stage, the performance of an actuator applied to an undercarriage of an aircraft, also called an actuator cylinder, is often tested through a load testing machine, for example, a spring is used for restraining a rear flight control mechanical control system of the actuator, which is disclosed in chinese patent CN203350054U, so that a test for accurately loading a test piece in a movement process is realized, and the movement process of the rear flight control mechanical control system of the actuator of an aircraft in a flight state is simulated. Particularly, in the performance test of an actuator by using an electromechanical system, the rigidity test is realized by a motor and a lead screw mechanism, for example, the rigidity test device and method of the electromechanical actuator disclosed in the patent CN107727341, a lead screw II and a lead screw I are arranged symmetrically, two sides of a tested piece are provided, and a symmetric manual loading mode is adopted, so that the damage to the tested electromechanical actuator caused by too large active loading control force is avoided, a laser sensor I and a laser sensor II with high displacement measurement precision are used, and the laser sensor I and the laser sensor II are placed on a parallel line, so that the measurement precision is higher, the moment arm loading force is realized by using the lead screw II and the lead screw I of a T-shaped large-pitch thread with a self-locking function, matching the lead screw connecting gear I, the lead screw connecting gear II, a stress gear and a stress gear, and the loading force is more accurate by directly connecting a tension pressure sensor in series with the tested electromechanical actuator.
The existing performance testing or testing device mainly aims at the design of an actuator and aims at testing the performance of the loaded actuator, but after actual service, for example, the existing performance testing or testing device is applied to an actuator cylinder for lifting a flight undercarriage, because the existing performance testing or testing device is comprehensively influenced by various factors such as airplane load, ground friction, acceleration, service time and the like in the service process, the strength and the rigidity of the existing performance testing or testing device can be changed, even sudden severe changes are brought, and the results of performance tests such as rigidity test, durability test, unstable load and the like in the design stage are no longer suitable at this moment, and the existing performance testing or testing device is difficult to be actually dismounted for testing and early warning.
Disclosure of Invention
According to a first aspect of the present invention, there is provided a method of estimating fatigue of an actuator of a flight landing gear, comprising:
acquiring a service history sample set of a flight undercarriage actuator, wherein the service history sample set comprises service data of the actuator and service data of a main shock absorber of an undercarriage; the service data of the actuator comprise external force load, ground friction force, environment temperature, hydraulic medium temperature, ground falling instant speed, acceleration, piston displacement, design rigidity, service time and service state; the service data of the main shock absorber of the undercarriage comprises design load, recovery coefficient, service time and service state of the main shock absorber;
preprocessing the service data of the actuator and the service data of the main shock absorber of the undercarriage, and marking the service state;
taking a preprocessed service history sample set as an input sample, taking a force load, a ground friction force, an environment temperature, a hydraulic medium temperature, a falling instant speed, an acceleration, a piston displacement, a design rigidity, a main shock absorber design load, a recovery coefficient, service time and a service state as characteristic vectors, and respectively training by adopting a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) to obtain a plurality of actuator fatigue estimation models, wherein the model output is an actuator fault mark;
counting the prediction precision value of each model in a preset service time period;
weighting and summing the prediction precision value of each model in a preset service time period to determine weighted prediction precision;
obtaining a model with the highest weighted prediction precision value as a final fatigue estimation prediction model;
and obtaining a characteristic vector of the aircraft as an input, inputting the characteristic vector into the fatigue estimation prediction model, and outputting a predicted actuator fault sign.
According to a second aspect of the present invention there is provided a flight landing gear actuator fatigue estimation system comprising:
one or more processors;
a memory storing instructions that are operable, which when executed by the one or more processors, cause the one or more processors to perform operations comprising the aforementioned flow of a method of flight landing gear actuator fatigue estimation.
According to a third aspect of the present invention there is provided a computer readable medium storing software comprising instructions executable by one or more computers, the instructions when executed causing the one or more computers to perform operations comprising the flow of the aforementioned flight landing gear actuator fatigue estimation method.
The invention provides a fatigue estimation method and system for an actuator of a flight undercarriage, which aim to screen out an optimal model under the condition of different service years through the service data of the actuator and the service data of a main shock absorber of the undercarriage and a multi-model fusion training mode. Due to the characteristics of heterogeneous data diversity and nonlinear variation of the service data of the actuator and the service data of the landing gear main shock absorber, the optimized regression under multi-sample and multi-characteristic parameters is realized through the nonlinear mapping capability and the learning speed performance of the GRNN, the convergence speed is high, and the radial basis network is taken as the basis, so that the landing gear main shock absorber has good nonlinear approximation performance; the RBF neural network is a typical radial propagation neural network, a hidden layer space is formed by using RBF (radial basis function) as a 'base' of a hidden unit, the hidden layer transforms an input vector and transforms low-dimensional mode input data into a high-dimensional space, so that the problem of inseparability of linearity in the low-dimensional space is linearly separable in the high-dimensional space, the training time is shortened and the training progress is greatly controllable when the characteristics of multi-dimensional data of the actuator and the multi-dimensional data of the main shock absorber are input; in the DNN network structure, the chain forward propagation of the DNN deep neural network is utilized, and the training and prediction output on the basis of mass characteristics is realized through full connection among a plurality of layers.
On the basis, the influence brought by the combination of the main shock absorber and the service time is further excavated by combining with the actuator cylinder form and the design performance adopted by different aircrafts, an optimal prediction model is screened out through comprehensive judgment of weighted prediction precision, the fatigue state, namely the fault characteristic, of the actuator of the aircraft of the same model after service is analyzed and predicted, early warning and warning are realized, and the defect of subjective judgment brought by only depending on cabin ground inspection and visual observation is avoided.
It should be understood that all combinations of the foregoing concepts and additional concepts described in greater detail below can be considered as part of the inventive subject matter of this disclosure unless such concepts are mutually inconsistent. Additionally, all combinations of claimed subject matter are considered a part of the presently disclosed subject matter.
The foregoing and other aspects, embodiments and features of the present teachings can be more fully understood from the following description taken in conjunction with the accompanying drawings. Additional aspects of the present invention, such as features and/or advantages of exemplary embodiments, will be apparent from the description which follows, or may be learned by practice of the specific embodiments according to the teachings of the present invention.
Drawings
The figures are not intended to be drawn to scale. In the drawings, each identical or nearly identical component that is illustrated in various figures may be represented by a like numeral. For purposes of clarity, not every component may be labeled in every drawing. Embodiments of various aspects of the present invention will now be described, by way of example, with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart illustrating a method for estimating fatigue of an actuator of a flight landing gear according to an embodiment of the invention.
FIG. 2 is a schematic diagram of a flight landing gear actuator fatigue estimation device according to an embodiment of the invention.
Detailed Description
In order to better understand the technical content of the present invention, specific embodiments are described below with reference to the accompanying drawings.
In this disclosure, aspects of the present invention are described with reference to the accompanying drawings, in which a number of illustrative embodiments are shown. Embodiments of the present disclosure are not necessarily intended to include all aspects of the invention. It should be appreciated that the various concepts and embodiments described above, as well as those described in greater detail below, may be implemented in any of numerous ways, as the disclosed concepts and embodiments are not limited to any one implementation. In addition, some aspects of the present disclosure may be used alone, or in any suitable combination with other aspects of the present disclosure.
With reference to the method for estimating fatigue of the actuator of the aircraft landing gear in the exemplary embodiment shown in fig. 1, an optimal prediction model is screened out by adopting comprehensive judgment based on multi-model fusion training and combined with weighted prediction precision, and the optimal prediction model is used for fault early warning and warning of the actuator of the aircraft landing gear in a service state, so that the defects of manual judgment and subjective judgment are reduced.
In connection with the exemplary flight landing gear actuator fatigue estimation method shown in fig. 1, a flowchart (S100) implementing the method includes the following processes:
s101: acquiring a service history sample set of a flight undercarriage actuator, wherein the service history sample set comprises actuator service data and undercarriage main shock absorber service data; the service data of the actuator comprise external force load, ground friction force, environment temperature, hydraulic medium temperature, ground falling instant speed, acceleration, piston displacement, design rigidity, service time and service state; the service data of the main shock absorber of the undercarriage comprises design load, recovery coefficient, service time and service state of the main shock absorber;
s102: preprocessing the service data of the actuator and the service data of the main shock absorber of the undercarriage, and marking the service state; the fault mark in the normal service state is 0, and the fault mark in the abnormal service state is 1;
s103: taking a preprocessed service history sample set as an input sample, taking a force load, a ground friction force, an environment temperature, a hydraulic medium temperature, a falling instant speed, an acceleration, a piston displacement, a design rigidity, a main shock absorber design load, a recovery coefficient, service time and a service state as characteristic vectors, and respectively training by adopting a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) to obtain a plurality of actuator fatigue estimation models, wherein the model output is an actuator fault mark;
s104: counting the prediction precision value of each model in a preset service time period;
s105: weighting and summing the prediction precision value of each model in a preset service time period to determine weighted prediction precision;
s106: obtaining a model with the highest weighted prediction precision as a final fatigue estimation prediction model;
s107: and obtaining a characteristic vector of the aircraft as an input, inputting the characteristic vector into the fatigue estimation prediction model, and outputting a predicted actuator fault sign.
In an embodiment of the present invention, the input samples are divided into a training set and a test set, wherein the training set is used for training the model, and the test set is used for testing the accuracy of model prediction. The prediction accuracy value is obtained by model prediction based on a test set.
For a certain model of aircraft, the service history sample set is obtained from the historical flight data of the aircraft of the model; and, the input sample includes more than 5000 sample data, at least 80% of them is used for training the model. In the embodiment of the present invention, the training set is 4500 sample data, and the test set is 500 sample data.
In the training process of step S103, taking the GRNN network as an example, the training process may adopt an existing training process, and an optional example training process includes:
firstly, defining the sizes of an input layer, a hidden layer, a summation layer and an output layer, constructing a GRNN model, determining an optimal smooth parameter by using a cross validation method, and training the GRNN;
then, based on the trained GRNN, the test set data is used as input to perform prediction output, and whether the module is suitable for the test set data is determined according to the judgment of the output result and the actual result. And if the accuracy of the prediction result is not satisfactory, retraining or optimizing the GRNN model until obtaining a GRNN neural network prediction model with a prediction error meeting the expected requirement.
Wherein, the GRNN network follows the following formula,
wherein X = [ X ] 1 ，x 2 ，x 3 ，...，x m ] T For the network input, i.e., the feature vector, y is the corresponding prediction output, representing the prediction output of y with the input being X, and f (X, y) is the joint probability density function of X and y.
In an alternative embodiment, the DNN and RBF network based training process may be implemented in conjunction with existing commercial training processes.
Therefore, model training of three network structures is carried out on the basis of the same training set, and corresponding prediction models are obtained.
It should be understood that a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) are used for training, and in the model training process, a corresponding prediction model is obtained by optimizing the prediction error judgment of a prediction result.
Further, on the basis of obtaining three models through training, the test sets can be respectively input, and corresponding prediction output, namely the actuator fault mark, is 1 or 0. And based on this output, comparing with actual values in the test set to determine the predicted output accuracy of each model.
In the embodiment of the present invention, in step S104, the method for calculating the prediction accuracy value of each model in the preset service time period specifically includes the following steps:
taking the service years as a period, counting the prediction accuracy value of each actuator fatigue estimation model under the corresponding service years, and expressing as follows: the prediction precision of the service years = the number of fault signs accurately predicted corresponding to the service years/the number of samples corresponding to the service years;
the cycle interval of the service years is set to 1 year, for example, 1 year, 2 years, 3 years, 4 years, 5 years, etc. of service.
In the embodiment of the present invention, in step S105, the weighted summation is performed on the prediction precision value of each model in the preset service time period, and the weighted prediction precision is determined, which specifically includes the following processes:
the weighted prediction accuracy for each model is calculated as follows:
f(1)＝w1*q11+w2*q12+w3*q13+...+wn*q1n
f(2)＝w1*q21+w2*q22+w3*q23+...+wn*q2n
f(3)＝w1*q31+w2*q32+w3*q33+...+wn*q3n
defining actuator fatigue estimation models obtained based on Generalized Regression Neural Network (GRNN), radial basis function neural network (RBF) and deep learning neural network (DNN) training as a first estimation model, a second estimation model and a third estimation model respectively; f (1), f (2) and f (3) respectively represent weighted prediction accuracy of the first estimation model, the second estimation model and the third estimation model;
q1n represents a prediction accuracy value of the first estimation model in the nth service year;
q2n represents a prediction precision value of the second estimation model in the nth service year;
q3n represents a prediction accuracy value of the third estimation model in the nth service year;
w1, w2, w3, \8230;, wn, which respectively represent weighted values of prediction accuracy values for different service years.
Wherein the weighted values of the prediction accuracy values of different service years are set as:
w1+w2+w3+…+wn＝1；
and w1> w2> w3> \8230wn.
The shorter the service life is, the closer the performance of the actuator and the main shock absorber is to the design standard, and the higher the required prediction accuracy is, so that in the process of performing weighting calculation when the prediction accuracy of different estimation models is screened, the higher the prediction accuracy is required for the actuator with the shorter service life is, and the higher the weight is.
For example, in an embodiment of the present invention, the data of the training set and the test set thereof is defined as follows:
training set
m1(x11,x12,x13,x14,x15,x16,x17,x18,x19,x110,x111,x112,mark1)
m2(x21,x22,x23,x24,x25,x26,x27,x28,x29,x210,x211,x212,mark2)
m3(x31,x32,x33,x34,x35,x36,x37,x38,x39,x310,x311,x312,mark3)
.
.
.
mn(xn1,xn2,xn3,xn4,xn5,xn6,xn7,xn8,xn9,xn10,xn11,xn12,markn)
The system comprises a main shock absorber, a plurality of sample data storage units, a plurality of characteristic vectors, an n1 sample, an n2 sample, an n3 sample, an n4 sample, an n5 sample, an n6 sample, an n7 sample, an n8 sample, an n9 sample, an n10 sample, an n11 sample, an n12 sample, an external load sample, a ground friction sample, an environmental temperature sample, a hydraulic medium temperature sample, a ground instantaneous speed sample, an acceleration sample, a piston displacement sample, a design rigidity sample, a main shock absorber design load sample, a recovery coefficient sample, service time sample and a service state sample.
Markn denotes a failure flag of the nth sample data.
The three prediction models generated by training are respectively input into a test set, the models predict and output corresponding fault signs, and then the operation is carried out in a calculation mode of the prediction precision of the service year to obtain the prediction precision value of each model in a preset service time period, wherein the example is as follows:
| first estimated model accuracy | Second estimated model accuracy | Third estimated model accuracy | |
| Service life of 1 year | q11 | q21 | q31 |
| Service life of 2 years | q12 | q22 | q32 |
| Service life of 3 years | q13 | q23 | q33 |
| ... | ... | ... | ... |
| Service life of n years | q1n | q2n | q3n |
With reference to fig. 2, a fatigue estimation device 200 for an actuator of a flight landing gear according to an embodiment of the present invention includes a service history sample set acquisition module, a service history sample set preprocessing module, a training module, a prediction precision value statistics module, a weighted prediction precision acquisition module, a fatigue estimation prediction model screening module, and a prediction output module.
The device comprises a service history sample set acquisition module, a service history sample set acquisition module and a service data acquisition module, wherein the service history sample set acquisition module is used for acquiring a service history sample set of a flight undercarriage actuator and comprises actuator service data and undercarriage main shock absorber service data. The service data of the actuator comprises external force load, ground friction force, environment temperature, hydraulic medium temperature, ground falling instant speed, acceleration, piston displacement, design rigidity, service time and service state; the service data of the main shock absorber of the landing gear comprises the design load, the recovery coefficient, the service time and the service state of the main shock absorber.
The system comprises a service history sample set preprocessing module, a service state detection module and a service state identification module, wherein the service history sample set preprocessing module is used for preprocessing service data of an actuator and service data of a main shock absorber of an undercarriage and marking a service state; wherein, the fault sign of the service state is normal is 0, and the fault sign of the service state is abnormal is 1.
The training module is used for training by taking the preprocessed service history sample set as input samples and respectively adopting a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) to obtain a plurality of actuator fatigue estimation models by taking an external force load, a ground friction force, an environment temperature, a hydraulic medium temperature, a ground instantaneous speed, an acceleration, a piston displacement, a design rigidity, a main shock absorber design load, a recovery coefficient, service time and a service state as characteristic vectors, wherein the output of each model is an actuator fault mark.
And the prediction accuracy value statistic module is used for counting the prediction accuracy value of each model in a preset service time period.
And the weighted prediction precision acquisition module is used for carrying out weighted summation on the prediction precision value of each model in a preset service time period and determining the weighted prediction precision.
And the fatigue estimation prediction model screening module is used for acquiring a model with the highest weighted prediction precision as a final fatigue estimation prediction model.
And the prediction output module is used for obtaining the characteristic vector of the aircraft as input, inputting the characteristic vector into the fatigue estimation prediction model and outputting the predicted actuator fault sign.
There is also provided, in accordance with an embodiment of the present invention, a flight landing gear actuator fatigue estimation system, including: one or more processors; a memory storing instructions that are operable, when executed by the one or more processors, to cause the one or more processors to perform operations comprising the flow of the flight landing gear actuator fatigue estimation method of the foregoing embodiments.
According to an embodiment of the invention, there is also provided a computer-readable medium storing software comprising instructions executable by one or more computers, the instructions, when executed, causing the one or more computers to perform operations comprising the flow of the method of estimating fatigue of a flight landing gear actuator of the previous embodiment.
Although the present invention has been described with reference to the preferred embodiments, it is not intended to be limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention. Therefore, the protection scope of the present invention should be defined by the appended claims.
Claims (10)
1. A method of estimating fatigue of an actuator of a flight landing gear, comprising:
acquiring a service history sample set of a flight undercarriage actuator, wherein the service history sample set comprises actuator service data and undercarriage main shock absorber service data; the service data of the actuator comprise external force load, ground friction force, environment temperature, hydraulic medium temperature, ground falling instant speed, acceleration, piston displacement, design rigidity, service time and service state; the service data of the main shock absorber of the undercarriage comprises design load, recovery coefficient, service time and service state of the main shock absorber;
preprocessing the service data of the actuator and the service data of the main shock absorber of the undercarriage, and marking the service state;
taking a preprocessed service history sample set as an input sample, taking a force load, a ground friction force, an environment temperature, a hydraulic medium temperature, a falling instant speed, an acceleration, a piston displacement, a design rigidity, a main shock absorber design load, a recovery coefficient, service time and a service state as characteristic vectors, and respectively training by adopting a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) to obtain a plurality of actuator fatigue estimation models, wherein the model output is an actuator fault mark;
counting the prediction precision value of each model in a preset service time period;
carrying out weighted summation on the prediction precision value of each model in a preset service time period, and determining weighted prediction precision;
obtaining a model with the highest weighted prediction precision as a final fatigue estimation prediction model;
and obtaining a characteristic vector of the aircraft as an input, inputting the characteristic vector into the fatigue estimation prediction model, and outputting a predicted actuator fault sign.
2. The method of estimating fatigue of an actuator of a flight landing gear according to claim 1, wherein the marking of the in-service condition is performed with a normal in-service fault flag of 0 and an abnormal in-service fault flag of 1.
3. The method for estimating fatigue of an actuator of a flight landing gear according to claim 1, wherein a Generalized Regression Neural Network (GRNN), a radial basis function neural network (RBF) and a deep learning neural network (DNN) are used for training, and in a model training process, a corresponding prediction model is obtained by optimizing prediction error judgment of a prediction result.
4. The method of estimating fatigue in an actuator of a flight landing gear according to claim 1, wherein the step of counting the predicted accuracy value of each model over a predetermined time period of service comprises:
taking the service years as a period, counting the prediction precision value of each actuator fatigue estimation model under the corresponding service years, and expressing as follows: the prediction precision of the service year = the number of fault signs predicted accurately corresponding to the service year/the number of samples corresponding to the service year;
wherein, the period interval of the service years is set as 1 year.
5. The method of estimating fatigue of an actuator of a flight landing gear according to claim 1, wherein the weighted summation of the prediction accuracy values of each model over a preset time period of service to determine the weighted prediction accuracy comprises:
the weighted prediction accuracy for each model is calculated in the following manner:
f(1)＝w1*q11+w2*q12+w3*q13+…+wn*q1n
f(2)＝w1*q21+w2*q22+w3*q23+…+wn*q2n
f(3)＝w1*q31+w2*q32+w3*q33+…+wn*q3n
defining actuator fatigue estimation models obtained based on Generalized Regression Neural Network (GRNN), radial basis function neural network (RBF) and deep learning neural network (DNN) training as a first estimation model, a second estimation model and a third estimation model respectively; f (1), f (2) and f (3) respectively represent weighted prediction accuracy of the first estimation model, the second estimation model and the third estimation model;
q1n represents a prediction accuracy value of the first estimation model in the nth service year;
q2n represents a prediction precision value of the second estimation model in the nth service year;
q3n represents a prediction accuracy value of the third estimation model in the nth service year;
w1, w2, w3, \ 8230and wn, which respectively represent weighted values of prediction accuracy values of different service years.
6. The method of estimating fatigue of an actuator of a flight landing gear according to claim 5, wherein the weights for the prediction accuracy values for different years of service are set to:
w1+w2+w3+…+wn＝1；
and w1> w2> w3> \8230wn.
7. A method of estimating fatigue of an actuator of a flight landing gear according to any one of claims 1 to 6, wherein the input samples are divided into a training set and a test set, wherein the training set is used for training a model and the test set is used for accuracy testing of model prediction;
wherein, the prediction precision value is obtained by model prediction based on the test set.
8. The method of estimating fatigue of an actuator of a flight landing gear according to claim 7, wherein for a model of aircraft, the service history sample set is obtained from historical flight data for the model of aircraft; and, the input sample includes more than 5000 sample data, at least 80% of them is used for training the model.
9. A flight landing gear actuator fatigue estimation system, comprising:
one or more processors;
a memory storing instructions that are operable, which when executed by the one or more processors, cause the one or more processors to perform operations comprising a flow of a flight landing gear actuator fatigue estimation method according to any one of claims 1 to 8.
10. A computer-readable medium storing software, wherein the software includes instructions executable by one or more computers which, when executed, cause the one or more computers to perform operations comprising a flow of a method of estimating fatigue of an actuator of a flight landing gear according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211049584.2A CN115422835A (en) | 2022-08-30 | 2022-08-30 | Flight landing gear actuator fatigue estimation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211049584.2A CN115422835A (en) | 2022-08-30 | 2022-08-30 | Flight landing gear actuator fatigue estimation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115422835A true CN115422835A (en) | 2022-12-02 |
Family
ID=84200468
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211049584.2A Pending CN115422835A (en) | 2022-08-30 | 2022-08-30 | Flight landing gear actuator fatigue estimation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115422835A (en) |
-
2022
- 2022-08-30 CN CN202211049584.2A patent/CN115422835A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112268799B (en) | Static strength and fatigue strength integrated test verification method for composite material structure | |
| CN112784356B (en) | Test method for static strength of airplane structure | |
| Baranowski et al. | Assessment of mechanical properties of offroad vehicle tire: Coupons testing and FE model development | |
| CN103942418A (en) | Method for determining specific fatigue strength rated value under multi-axial loading condition | |
| CN111581763B (en) | Method for evaluating air path fault diagnosis result of aero-engine | |
| CN115575104A (en) | Fan blade damage rapid detection method based on inverse finite element reconstruction image recognition | |
| CN110887731A (en) | Method and system for identifying internal flaws in a part produced using additive manufacturing | |
| EP2957883A2 (en) | Method for prognostics of an aircraft structure based on structural testing | |
| US20220067602A1 (en) | Management of risks related to the lack of compliance with a dimensional tolerance in a tolerance chain | |
| CN111191391A (en) | Simulation piece design method based on local stress distribution weighting coefficient | |
| CN112052523B (en) | Helicopter rotor moving part strength design method based on load characteristic comparison | |
| CN117669033A (en) | Single machine key part stress monitoring method based on flight parameter and bridge data | |
| Venugopal et al. | Fatigue Crack Growth Prediction on Su-30MKM Horizontal Stabilizer Lug Using Static Analysis | |
| CN115422835A (en) | Flight landing gear actuator fatigue estimation method and system | |
| RU2717750C1 (en) | Method of strength tests of natural structures | |
| CN114254533B (en) | Method for examining influence and prediction of fatigue vibration on fixed angle of product group component | |
| CN115048726A (en) | Method and system for calculating maximum stress distribution of vehicle chassis parts and storage medium | |
| US20220214244A1 (en) | Systems and methods for assessing structural health | |
| CN115081116A (en) | Method and system for calculating maximum stress of vehicle chassis part and storage medium | |
| CN105068529B (en) | A kind of fault separating method and system compared based on weighted residual component | |
| CN106644782A (en) | Method for predicting multi-axial fatigue crack initiation direction based on power spectrum density | |
| CN113033632A (en) | High-altitude platform fault diagnosis method based on wavelet analysis and multi-layer overrun learning machine | |
| CN118503609A (en) | Flight load measurement method, system and medium based on multiple linear regression theory | |
| d'Ippolito et al. | Reliability-based design optimization of composite and steel aerospace structures | |
| Tikka | Fatigue life evaluation of critical locations in aircraft structures using virtual fatigue test |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |