CN113049409B - Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy - Google Patents

Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy Download PDF

Info

Publication number
CN113049409B
CN113049409B CN202110158259.9A CN202110158259A CN113049409B CN 113049409 B CN113049409 B CN 113049409B CN 202110158259 A CN202110158259 A CN 202110158259A CN 113049409 B CN113049409 B CN 113049409B
Authority
CN
China
Prior art keywords
impact
wavelet
signal
frequency
flexible arm
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.)
Active
Application number
CN202110158259.9A
Other languages
Chinese (zh)
Other versions
CN113049409A (en
Inventor
曾捷
钟照振
冯振辉
吴亚星
岳应萍
徐云涛
李开宇
綦磊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202110158259.9A priority Critical patent/CN113049409B/en
Publication of CN113049409A publication Critical patent/CN113049409A/en
Application granted granted Critical
Publication of CN113049409B publication Critical patent/CN113049409B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
    • G01N3/30—Investigating strength properties of solid materials by application of mechanical stress by applying a single impulsive force, e.g. by falling weight
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N3/00—Investigating strength properties of solid materials by application of mechanical stress
    • G01N3/02—Details
    • G01N3/06—Special adaptations of indicating or recording means
    • G01N3/068—Special adaptations of indicating or recording means with optical indicating or recording means

Abstract

The invention discloses a thin-wall flexible arm structure optical fiber impactor identification method based on signal frequency band energy, and belongs to the technical field of structural health monitoring. The method comprises the following steps: the thin-wall composite material flexible arm structure optical fiber Bragg grating sensor network is arranged; step two: acquiring FBG sensor response signals of the thin-wall composite material flexible arm structure under the impact action of different types of impact objects; step three: removing the baseline interference in the FBG sensor response signal through a wavelet decomposition technology, and reconstructing an impact response signal; step four: performing wavelet packet transformation on the reconstructed impulse response signal to obtain signal frequency band energy distribution; step five: calculating a sample impact point frequency band energy characteristic quantity proportion parameter, and constructing a characteristic quantity sample library; step six: and determining the type of the impact object by taking the minimum error value of the band energy characteristic quantity of the test point and the Euclidean distance of the sample library as a criterion.

Description

Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy
Technical Field
The invention belongs to the technical field of load monitoring of structural health monitoring, and particularly provides a thin-wall flexible arm structure optical fiber impactor identification method based on signal frequency band energy.
Background
The composite material structure has been widely used in the aerospace field due to its high specific strength, high specific stiffness, and good fatigue resistance and durability. The composite material flexible arm structure has the advantages of light weight, high folding and unfolding repeatability precision, simple folding and unfolding principle and the like, and is widely applied to aerospace vehicles such as a plane array surface supporting frame, a flexible detection arm and the like. However, the composite flexible arm structure is vulnerable to impact from external objects such as bird strikes, space debris, runway flyrock, etc. during service, thereby damaging the structure. The damage caused by the impact on the flexible arm structure of the composite material is usually sudden, and particularly, some low-speed impacts with small energy are not marked on the surface of an object, but visually undetectable damage forms such as microcracks, delamination and fiber breakage can be caused inside the composite material structure, so that key mechanical performance indexes such as tensile strength and compressive strength of the composite material structure are obviously reduced. Therefore, in order to save the overhaul cost and improve the maintenance efficiency, the type of the impact object needs to be identified, and a countermeasure needs to be taken according to the type of the impact object to avoid the occurrence of an accident.
However, the domestic impact object identification research is basically in a blank state at present, a real and effective impact object identification method is not provided, and test verification is not carried out on a composite material flexible arm structure. Based on the analysis, the invention provides a thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy. The method is suitable for the field of health monitoring of thin-wall flexible arm structures, and can effectively identify the types and impact point positions of different impact objects.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the blank of the identification method of the impact object of the thin-wall flexible arm structure, the invention provides the identification method of the impact object for the composite material flexible arm structure. The method comprises the steps of adopting an FBG sensor network to sense impact load response signals at different positions in a structure, extracting frequency band energy characteristic quantity, constructing a sample library, and combining a corresponding algorithm to realize identification of impact object types and impact point positions.
The technical scheme is as follows: in order to achieve the purpose, the technical scheme adopted by the invention is as follows:
the method comprises the following steps: FBG sensor network arrangement of thin-wall composite material flexible arm structure
The thin-wall composite material flexible arm structure consists of two parts which are symmetrical up and down, a monitoring area is selected at the central part of the upper surface of the single-end fixed thin-wall composite material flexible arm structure, and the monitoring area is a rectangular area after being unfolded according to a plane; establishing a coordinate system by taking the structure extending direction as an X direction and the arc direction as a Y direction after being unfolded; a row of FBG sensors parallel to the structure extending direction is respectively arranged at two ends of the monitoring area at equal intervals, each row is respectively provided with N sensors (4< N <16), the FBG sensors close to the fixed end are a first row, and the FBG sensors close to the free end are a second row. The FBG sensors in each column are in a wavelength division multiplexing layout form, the FBG sensors in different columns are in a space division multiplexing layout form, and the FBG sensors are adhered to the upper surface of the test piece structure, so that a distributed sensor network is formed.
Step two: collecting FBG sensor response signals of thin-wall composite material flexible arm structure under impact action of different types of impact objects
In a monitoring area with a safety distance of 50mm from the FBG sensors at two ends, a plurality of lines parallel to the X direction are uniformly divided, a plurality of columns parallel to the Y direction are uniformly divided, and the intersection points of the X, Y direction straight lines form Q sample impact points. Respectively applying impact loads at Q sample impact points on the surface of the thin-wall composite material flexible arm structure by adopting a PCB impact force hammer, and realizing a scene that different impactors impact the thin-wall composite material flexible arm structure by replacing impactors with different elastic moduli; and respectively recording impact response signals of the FBG sensors acting on different sample impact points at different impact objects.
Step three: removing the baseline interference in the FBG sensor response signal through a wavelet decomposition technology and reconstructing the shock response signal
Step a: according to the wavelet decomposition theory, the wavelet analysis has the characteristic of variable resolution, the multi-scale refinement analysis of non-stationary impact response signals can be realized, and the time-frequency domain characteristics of the signals are extracted.
Arbitrary function f (t) e L2(R) continuous wavelet transform:
Figure BDA0002934669740000021
wherein L is2(R) represents a two-dimensional real number domain, R being a real number, #a,b(t) is a wavelet series. Where a is a scale factor, also called a scale parameter, which determines the wavelet psia,b(t) frequency domain center and bandwidth; b is called a shift factor, and the shift factor and the scale factor together determine the time domain center of the wavelet. The reconstruction formula of f (t) is:
Figure BDA0002934669740000022
by changing the expansion parameter a and the translation factor b, the projection widths of the function f (t) on different time-frequency widths can be changed to correspond to the original signal, so that the time-frequency domain local analysis of the signal is more detailed, and low-frequency and high-frequency detail coefficients are obtained.
Step b: and (b) performing wavelet decomposition on the impact response signal obtained in the step (II) through the step (a), selecting a proper wavelet function (db 1-db 10), and performing P-layer wavelet decomposition on the impact response signal (3< P <12), so as to obtain low-frequency and high-frequency detail coefficients.
Step c: and c, extracting low-frequency and high-frequency detail coefficients of each order obtained in the step b, setting the low-frequency detail coefficients to be zero, and reconstructing the residual high-frequency detail coefficients to obtain an impact response signal without baseline interference.
Step four: performing wavelet packet transformation on the reconstructed impulse response signal to obtain the frequency band energy distribution of the response signal
Step a: the wavelet packet is formed by linearly combining a series of wavelet functions, and belongs to L for any signal f (t) e2(R) can be decomposed into the following wavelet packets:
Figure BDA0002934669740000031
wherein
Figure BDA0002934669740000032
And
Figure BDA0002934669740000033
respectively wavelet packet coefficients and wavelet functions, the wavelet packet functions being obtainable by variations of the wavelet functions, the wavelet packet coefficients
Figure BDA0002934669740000034
By passing
Figure BDA0002934669740000035
Thus obtaining the product.
Step b: through the step a in the step four, the impulse response signal without the baseline interference is decomposed into a plurality of component signals with equal bandwidth, the corresponding window contains information of specific frequency spectrum and signal energy, and the energy of each component signal reflects the signal energy distribution of the original signal in each frequency band.
Is provided with
Figure BDA0002934669740000036
For wavelet component signals of wavelet packet decomposition, the i-th component signal at the j-scale has energy of:
Figure BDA0002934669740000037
where T represents the sampling time, the band energy distribution of the impulse response signal with the baseline disturbance removed is obtained by equation (4).
The specific method comprises the following steps: selecting proper wavelet basis function to carry out P-layer wavelet packet decomposition on the signal to obtain 2^PThe order component energy signals and the amplitude energy of each order signal are calculated; and finally, normalizing the amplitude energy of each order of signal.
Step five: calculating the proportional parameters of the energy characteristic quantities of the sample impact points and the frequency bands, and constructing a characteristic quantity sample library
Step a: defining a pre-division T (1)<T<2^P) Sum E of band energies of all remaining component signals excepthThe sum of the total band energy EsumProportional parameter R ofEThen, the calculation formula is:
Figure BDA0002934669740000041
wherein R isE,iCalculating the characteristic quantity proportional parameter for the ith FBG sensor, El,iRepresenting the band energy of the pre-T order component signal of the ith FBG sensor, Esum,iThe band energy of all component signals for the ith FBG sensor.
Step b: assuming that M impact objects exist, the proportion parameter sample library of each type of impact object is named as R1、R2、…、RMThe composition is represented as:
Figure BDA0002934669740000042
in the equation (6), a indicates that an impact load is applied at the impact point a, i indicates the ith FBG sensor, and j indicates the jth impact object. The proportion parameter sample library R of M impact objects at the impact point QE,QThe collection is as follows:
RE,Q＝[R1,Q,R2,Q,…,Rj,Q] (j＝1,2,…,M) (7)
for Q grid impact points, R of M impact objects at all impact pointsEEstablishing a proportion parameter sample matrix R according to the sequence of impact pointsE,ref，RE,refCan be expressed as:
Figure BDA0002934669740000043
in (8), Q represents that an impact load is applied at the impact point Q.
Step six: determining the type of the impact object by taking the minimum Euclidean distance error value of the ratio parameter of the band energy characteristic quantity of the test point and the ratio parameter of the sample library as a criterion
Defining a target function of Euclidean distance Error values of a test point frequency band energy characteristic quantity proportional parameter and a sample base proportional parameter as Error (R)E) The functional form is:
Figure BDA0002934669740000051
in (9), n is the number of FBG sensors.
For the impact object acting on a certain test point, when the frequency band energy characteristic quantity proportional parameters of the impact response signals measured by the 2N FBG sensors and Euclidean distance error values of all proportional parameters of the sample library are calculated, the method for judging the types of the impact point and the impact object is as follows:
(1) extracting an objective function Error (R)E) Minimum value of (c) min | Error (R)E)|i,j
(2) For min | Error (R) in (1)E)|i,jIf i is 1, the impact point position is 1. Similarly, i is 1,2, …, Q, and the impact point positions corresponding in sequence are 1,2, …, Q.
(3) For min | Error (R) in (1)E)|i,jIf j ∈ [1,2, …, 2N)]If the type of the impact object is the type 1 impact object; if j is equal to [2N +1,2N +2, …,4N ∈]If the type of the impact object is the type 2 impact object; if j is ∈ [2 Nx (M-1),2 Nx (M-1) +1, …,2 MxN]And the type of the impact object is the j-th type impact object.
Has the advantages that: the invention provides a thin-wall flexible arm structure optical fiber impactor identification method based on signal frequency band energy. The method is suitable for the field of health monitoring of thin-wall flexible arm structures, and can effectively identify the types and impact point positions of different impact objects.
Drawings
FIG. 1 is a diagram of a thin-walled composite material flexible arm structure FBG sensor network layout;
fig. 2 is a flow chart of impact recognition.
Detailed Description
The technical solution of the present invention is further described in detail below with reference to the accompanying drawings.
The method comprises the following steps: FBG sensor network arrangement of thin-wall composite material flexible arm structure
The thin-wall composite material flexible arm structure consists of two parts which are symmetrical up and down, a monitoring area is selected at the central part of the upper surface of the single-end fixed thin-wall composite material flexible arm structure, and the monitoring area is a rectangular area after being unfolded according to a plane; establishing a coordinate system by taking the structure extending direction as an X direction and the arc direction as a Y direction after being unfolded; a row of FBG sensors parallel to the structure extending direction is respectively arranged at two ends of the monitoring area at equal intervals, each row is respectively provided with N sensors (4< N <16), the FBG sensors close to the fixed end are a first row, and the FBG sensors close to the free end are a second row. The FBG sensors in each row are arranged in a wavelength division multiplexing manner, and the FBG sensors in different rows are arranged in a space division multiplexing manner, and the FBG sensors are adhered to the upper surface of the test piece structure, so that a distributed sensor network is formed, as shown in fig. 1.
Step two: collecting FBG sensor response signals of thin-wall composite material flexible arm structure under impact action of different types of impact objects
In a monitoring area with a safety distance of 50mm from the FBG sensors at two ends, five lines parallel to the X direction are uniformly divided, five lines parallel to the Y direction are uniformly divided, and intersection points of X, Y direction lines form 25 sample impact points which are respectively represented by A-Y. Applying impact loads to 25 sample impact points on the surface of the thin-wall composite material flexible arm structure by adopting a PCB impact force hammer, and realizing a scene that different impactors impact the thin-wall composite material flexible arm structure by replacing impactors with different elastic moduli; and respectively recording impact response signals of the FBG sensors acting on different sample impact points at different impact objects.
Step three: removing the baseline interference in the FBG sensor response signal through a wavelet decomposition technology and reconstructing the shock response signal
Step a: according to the wavelet decomposition theory, the wavelet analysis has the characteristic of variable resolution, the multi-scale refinement analysis of non-stationary impact response signals can be realized, and the time-frequency domain characteristics of the signals are extracted.
Arbitrary function f (t) e L2(R) continuous wavelet transform:
Figure BDA0002934669740000061
wherein psia,b(t) is a wavelet series. Where a is a scale factor, also called a scale parameter, which determines the wavelet psia,b(t) frequency domain center and bandwidth; b is called a shift factor, and the shift factor and the scale factor together determine the time domain center of the wavelet. The reconstruction formula of f (t) is:
Figure BDA0002934669740000062
by changing the expansion parameter a and the translation factor b, the projection widths of the function f (t) on different time-frequency widths can be changed to correspond to the original signal, so that the time-frequency domain local analysis of the signal is more detailed, and low-frequency and high-frequency detail coefficients are obtained.
Step b: and (c) performing wavelet decomposition on the impact response signal obtained in the step (II) through the step (a), selecting a db10 wavelet function, and performing 6-layer wavelet decomposition on the impact response signal to obtain low-frequency and high-frequency detail coefficients.
Step c: and c, extracting low-frequency and high-frequency detail coefficients of each order obtained in the step b, setting the low-frequency detail coefficients to be zero, and reconstructing the residual high-frequency detail coefficients to obtain an impact response signal without baseline interference.
Step four: performing wavelet packet transformation on the reconstructed impulse response signal to obtain the frequency band energy distribution of the response signal
Step a: the wavelet packet is formed by linearly combining a series of wavelet functions, and belongs to L for any signal f (t) e2(R) can be decomposed into the following wavelet packets:
Figure BDA0002934669740000071
wherein
Figure BDA0002934669740000072
And
Figure BDA0002934669740000073
respectively wavelet packet coefficients and wavelet functions, the wavelet packet functions being obtainable by variations of the wavelet functions, the wavelet packet coefficients
Figure BDA0002934669740000074
By passing
Figure BDA0002934669740000075
Thus obtaining the product.
Step b: through the step a in the step four, the impulse response signal without the baseline interference is decomposed into a plurality of component signals with equal bandwidth, the corresponding window contains information of specific frequency spectrum and signal energy, and the energy of each component signal reflects the signal energy distribution of the original signal in each frequency band.
Is provided with
Figure BDA0002934669740000076
For wavelet component signals of wavelet packet decomposition, the i-th component signal at the j-scale has energy of:
Figure BDA0002934669740000077
where T represents the sampling time, the band energy distribution of the impulse response signal with the baseline disturbance removed is obtained by equation (4).
The specific method comprises the following steps: selecting db10 wavelet basis functions to carry out 6-layer wavelet packet decomposition on the signals to obtain 64-order component energy signals, and calculating amplitude energy of each order of signals; and finally, normalizing the amplitude energy of each order of signal.
Step five: calculating the proportional parameters of the energy characteristic quantities of the sample impact points and the frequency bands, and constructing a characteristic quantity sample library
Step a: defining a sum E of the band energies of all remaining component signals except the first 4 th orderhThe sum of the total band energy EsumProportional parameter R ofEThen, the calculation formula is:
Figure BDA0002934669740000078
wherein R isE,iCalculating the characteristic quantity proportional parameter for the ith FBG sensor, El,iRepresenting the band energy of the first 4 th order component signal of the ith FBG sensor, Esum,iThe band energy of all component signals for the ith FBG sensor.
Step b: assuming that M impact objects exist, the proportion parameter sample library of each type of impact object is named as R1、R2、…、RMThe composition is represented as:
Figure BDA0002934669740000081
in the equation (6), a indicates that an impact load is applied at the impact point a, i indicates the ith FBG sensor, and j indicates the jth impact object. The proportional parameter sample library R of four impact objects at the impact point A is storedE,AThe collection is as follows:
RE,A＝[R1,A,R2,A,…,Rj,A] (j＝1,2,…,M) (7)
for 25 grid impact points, all impact points are R of four impact objectsEEstablishing a proportion parameter sample matrix R according to the sequence of impact pointsE,ref，RE,refCan be expressed as:
Figure BDA0002934669740000082
in (8), Y is indicated as an impact load applied at an impact point Y.
Step six: determining the type of the impact object by taking the minimum Euclidean distance error value of the ratio parameter of the band energy characteristic quantity of the test point and the ratio parameter of the sample library as a criterion
Defining a target function of Euclidean distance Error values of a test point frequency band energy characteristic quantity proportional parameter and a sample base proportional parameter as Error (R)E) The functional form is:
Figure BDA0002934669740000083
in (9), n is the number of FBG sensors.
For the impact object acting on a certain test point, when the frequency band energy characteristic quantity proportional parameters of the impact response signals measured by 10 FBG sensors and Euclidean distance error values of all proportional parameters of a sample library are calculated, the method for judging the types of the impact point and the impact object is as follows:
(1) extracting an objective function Error (R)E) Minimum value of (c) min | Error (R)E)|i,j
(2) For min | Error (R) in (1)E)|i,jIf i is 1, the impact point is a. Similarly, i is 1,2, …,25, and the impact point positions are A, B, …, Y, respectively.
(3) For min | Error (R) in (1)E)|i,jIf j ∈ [1,2, …,10 ]]If the type of the impact object is the type 1 impact object; if j ∈ [11,12, …,20]If the type of the impact object is the type 2 impact object; if j is ∈ [10 × (M-1),10 × (M-1) +1, …,10M]And the type of the impact object is the j-th type impact object.

Claims (1)

1. A thin-wall flexible arm structure optical fiber impactor identification method based on signal frequency band energy is characterized by comprising the following steps:
the method comprises the following steps: FBG sensor network arrangement of thin-wall composite material flexible arm structure
The thin-wall composite material flexible arm structure consists of two parts which are symmetrical up and down, a monitoring area is selected at the central part of the upper surface of the single-end fixed thin-wall composite material flexible arm structure, and the monitoring area is a rectangular area after being unfolded according to a plane; establishing a coordinate system by taking the structure extending direction as an X direction and the arc direction as a Y direction after being unfolded; arranging a row of FBG sensors parallel to the structure extending direction at equal intervals at two ends of the monitoring area respectively, wherein each row is provided with N sensors, 4< N <16, the FBG sensors close to the fixed end are a first row, and the FBG sensors close to the free end are a second row; the FBG sensors in each row are distributed in a wavelength division multiplexing mode, different rows are distributed in a space division multiplexing mode, and the FBG sensors are adhered to the upper surface of the test piece structure to form a distributed sensor network;
step two: collecting FBG sensor response signals of thin-wall composite material flexible arm structure under impact action of different types of impact objects
In a monitoring area with a safety distance of 50mm from the FBG sensors at two ends, uniformly dividing a plurality of lines of straight lines parallel to the X direction, uniformly dividing a plurality of columns of straight lines parallel to the Y direction, and forming Q sample impact points by the intersection points of the straight lines in the X, Y direction; respectively applying impact loads at Q sample impact points on the surface of the thin-wall composite material flexible arm structure by adopting a PCB impact force hammer, and realizing a scene that different impactors impact the thin-wall composite material flexible arm structure by replacing impactors with different elastic moduli; respectively recording impact response signals of the FBG sensors acting on different sample impact points at different impact objects;
step three: removing the baseline interference in the FBG sensor response signal through a wavelet decomposition technology and reconstructing the shock response signal
Step a: according to the wavelet decomposition theory, the wavelet analysis has the characteristic of variable resolution, the multi-scale refinement analysis of non-stable impact response signals can be realized, and the time-frequency domain characteristics of the signals are extracted;
arbitrary function f (t) e L2Continuous wavelet transform of (R)Comprises the following steps:
Figure FDA0003505899790000011
wherein L is2(R) represents a two-dimensional real number domain, R being a real number, #a,b(t) is a wavelet series; where a is a scale factor, also called a scale parameter, which determines the wavelet psia,b(t) frequency domain center and bandwidth; b is called a translation factor, and the translation factor and the scale factor jointly determine the time domain center of the wavelet; the reconstruction formula of f (t) is:
Figure FDA0003505899790000012
by changing the expansion parameter a and the translation factor b, the projection widths of the function f (t) on different time-frequency widths can be changed to correspond to the original signal, so that the time-frequency domain local analysis of the signal is more detailed, and low-frequency and high-frequency detail coefficients are obtained;
step b: performing wavelet decomposition on the impact response signal obtained in the step two through the step a, selecting proper wavelet functions db 1-db 10, and performing P-layer wavelet decomposition on the impact response signal to obtain low-frequency and high-frequency detail coefficients, wherein 3< P < 12;
step c: b, extracting low-frequency and high-frequency detail coefficients of each order obtained in the step b, setting the low-frequency detail coefficients to zero, and reconstructing the remaining high-frequency detail coefficients to obtain an impact response signal without baseline interference;
step four: performing wavelet packet transformation on the reconstructed impulse response signal to obtain the frequency band energy distribution of the response signal
Step a: the wavelet packet is formed by linearly combining a series of wavelet functions, and belongs to L for any signal f (t) e2(R) can be decomposed into the following wavelet packets:
Figure FDA0003505899790000021
wherein
Figure FDA0003505899790000022
And
Figure FDA0003505899790000023
respectively wavelet packet coefficients and wavelet functions, the wavelet packet functions being obtainable by variations of the wavelet functions, the wavelet packet coefficients
Figure FDA0003505899790000024
By passing
Figure FDA0003505899790000025
Obtaining;
step b: through the step a in the step four, the impulse response signal without the baseline interference is decomposed into a plurality of component signals with the same bandwidth, the corresponding window contains information of a specific frequency spectrum and signal energy, and the energy of each component signal reflects the signal energy distribution condition of the original signal in each frequency band;
is provided with
Figure FDA0003505899790000026
For wavelet component signals of wavelet packet decomposition, the i-th component signal at the j-scale has energy of:
Figure FDA0003505899790000027
wherein T represents sampling time, and the frequency band energy distribution of the shock response signal without the baseline interference is obtained by the formula (4);
the specific method comprises the following steps: selecting proper wavelet basis function to carry out P-layer wavelet packet decomposition on the signal to obtain 2^PThe order component energy signals and the amplitude energy of each order signal are calculated; finally, normalizing the amplitude energy of each order of signal;
step five: calculating the proportional parameters of the energy characteristic quantities of the sample impact points and the frequency bands, and constructing a characteristic quantity sample library
Step a: defining a sum E of the band energies of all remaining component signals except the first ThThe sum of the total band energy EsumProportional parameter R ofEIn which 1 is<T<2^PThen, the calculation formula is:
Figure FDA0003505899790000028
wherein R isE,iCalculating the characteristic quantity proportional parameter for the ith FBG sensor, El,iRepresenting the band energy of the pre-T order component signal of the ith FBG sensor, Esum,iThe band energy of all component signals of the ith FBG sensor;
step b: assuming that M impact objects exist, the proportion parameter sample library of each type of impact object is named as R1、R2、…、RMThe composition is represented as:
Figure FDA0003505899790000031
in the formula (6), A represents that an impact load is applied at an impact point A, i represents the ith FBG sensor, and j represents the jth class impact object; the proportion parameter sample library R of M impact objects at the impact point QE,QThe collection is as follows:
RE,Q＝[R1,Q,R2,Q,…,Rj,Q]where j is 1,2, …, M (7)
For Q grid impact points, R of M impact objects at all impact pointsEEstablishing a proportion parameter sample matrix R according to the sequence of impact pointsE,ref，RE,refCan be expressed as:
Figure FDA0003505899790000032
in (8), Q is expressed as applying an impact load at an impact point Q;
step six: determining the type of the impact object by taking the minimum Euclidean distance error value of the ratio parameter of the band energy characteristic quantity of the test point and the ratio parameter of the sample library as a criterion
Defining a target function of Euclidean distance Error values of a test point frequency band energy characteristic quantity proportional parameter and a sample base proportional parameter as Error (R)E) The functional form is:
Figure FDA0003505899790000033
wherein i is 1,2, …, 2N; j-1, 2, …, M
In the step (9), n is the number of FBG sensors;
for the impact object acting on a certain test point, when the frequency band energy characteristic quantity proportional parameters of the impact response signals measured by the 2N FBG sensors and Euclidean distance error values of all proportional parameters of the sample library are calculated, the method for judging the types of the impact point and the impact object is as follows:
(1) extracting an objective function Error (R)E) Minimum value of (c) min | Error (R)E)|i,j
(2) For min | Error (R) in (1)E)|i,jIf i is 1, the impact point position is 1; similarly, i is 1,2, …, Q, the impact point positions corresponding in order are 1,2, …, Q;
(3) for min | Error (R) in (1)E)|i,jIf j ∈ [1,2, …, 2N)]If the type of the impact object is the type 1 impact object; if j is equal to [2N +1,2N +2, …,4N ∈]If the type of the impact object is the type 2 impact object; if j is ∈ [2 Nx (M-1),2 Nx (M-1) +1, …,2 MxN]And the type of the impact object is the j-th type impact object.
CN202110158259.9A 2021-02-04 2021-02-04 Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy Active CN113049409B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110158259.9A CN113049409B (en) 2021-02-04 2021-02-04 Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110158259.9A CN113049409B (en) 2021-02-04 2021-02-04 Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy

Publications (2)

Publication Number Publication Date
CN113049409A CN113049409A (en) 2021-06-29
CN113049409B true CN113049409B (en) 2022-04-22

Family

ID=76508899

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110158259.9A Active CN113049409B (en) 2021-02-04 2021-02-04 Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy

Country Status (1)

Country Link
CN (1) CN113049409B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114754691B (en) * 2022-02-28 2023-03-31 南京航空航天大学 Distributed optical fiber monitoring and inversion method for helicopter blade bending form

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104019739A (en) * 2014-05-08 2014-09-03 南京航空航天大学 Plate structure impact load positioning method based on distributed fiber bragg grating sensing network
CN106528945A (en) * 2016-10-17 2017-03-22 南京航空航天大学 Loading included angle normalization-based plate structure load position and size identification method
CN106546319A (en) * 2016-10-17 2017-03-29 南京航空航天大学 Clamped plate Impact monitoring method based on fiber grating transient response mathematic interpolation
CN106643484A (en) * 2016-10-17 2017-05-10 南京航空航天大学 Impact positioning method based on transient fractal calculation and similar two-dimensional normal distribution fitting
CN107607303A (en) * 2017-09-13 2018-01-19 河海大学 Mechanical Failure of HV Circuit Breaker recognition methods based on wavelet packet Yu SOM networks
CN109903525A (en) * 2019-01-23 2019-06-18 昆明理工大学 A kind of Rock And Soil unstability disaster monitoring method for early warning based on external exciting and Wavelet Packet Energy Spectrum

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104019739A (en) * 2014-05-08 2014-09-03 南京航空航天大学 Plate structure impact load positioning method based on distributed fiber bragg grating sensing network
CN106528945A (en) * 2016-10-17 2017-03-22 南京航空航天大学 Loading included angle normalization-based plate structure load position and size identification method
CN106546319A (en) * 2016-10-17 2017-03-29 南京航空航天大学 Clamped plate Impact monitoring method based on fiber grating transient response mathematic interpolation
CN106643484A (en) * 2016-10-17 2017-05-10 南京航空航天大学 Impact positioning method based on transient fractal calculation and similar two-dimensional normal distribution fitting
CN107607303A (en) * 2017-09-13 2018-01-19 河海大学 Mechanical Failure of HV Circuit Breaker recognition methods based on wavelet packet Yu SOM networks
CN109903525A (en) * 2019-01-23 2019-06-18 昆明理工大学 A kind of Rock And Soil unstability disaster monitoring method for early warning based on external exciting and Wavelet Packet Energy Spectrum

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
柔性机械臂振动控制中的压电传感器故障诊断;马天兵等;《电子测量与仪器学报》;20141215(第12期);全文 *
转子故障信号的小波能量特征自动提取;邓堰等;《应用科学学报》;20070915(第05期);全文 *

Also Published As

Publication number Publication date
CN113049409A (en) 2021-06-29

Similar Documents

Publication Publication Date Title
CN108932480B (en) Distributed optical fiber sensing signal feature learning and classifying method based on 1D-CNN
CN103617684B (en) Interference-type optical fiber circumference vibrating intruding recognizer
Lakshmi et al. Singular spectrum analysis combined with ARMAX model for structural damage detection
Chandrashekhar et al. Structural damage detection using modal curvature and fuzzy logic
CN105424359A (en) Sparse-decomposition-based hybrid fault feature extraction method of gear wheel and bearing
CN112067116B (en) Method for testing and analyzing impact vibration of medium and small bridges with noise resistance
CN105760934A (en) Bridge abnormity monitoring restoration method based on wavelet and BP neural network
CN110985310B (en) Wind driven generator blade fault monitoring method and device based on acoustic sensor array
CN113049409B (en) Thin-wall flexible arm structure optical fiber impact object identification method based on signal frequency band energy
Mendrok et al. Experimental verification of the damage localization procedure based on modal filtering
CN109933872B (en) Reinforced sparsity-based composite material structure impact load identification method and device
Shi et al. Structural damage detection using artificial neural networks and wavelet transform
Min et al. Online condition diagnosis for a two-stage gearbox machinery of an aerospace utilization system using an ensemble multi-fault features indexing approach
CN115754022A (en) Steel truss welding node damage degree identification method based on acoustic emission
CN108051126A (en) A kind of Varying-thickness Composite Laminated Plate under Low-Velocity Impact Thin interbed system and its method of work
CN113392545B (en) Time domain distribution random dynamic load identification method considering spatial correlation
Wang et al. Identification of engine faults based on acoustic emission signals using a 1DCNN-ViT ensemble model
CN115597901B (en) Bridge expansion joint damage monitoring method
Kolakowski et al. Two approaches to structural damage identification: model updating versus soft computing
Siebel et al. Assessment of mode shape-based damage detection methods under real operational conditions
CN114692465B (en) Nondestructive identification method, storage medium and equipment for bridge damage position
CN107628268A (en) Unilateral clamped wing Impact Location Method based on low frequency Coefficients of Approximation amplitude of variation
CN113049410B (en) Composite material laminate optical fiber impact position identification method based on strain nonlinear weighting
Yu et al. Impact localization system for composite barrel structure using fiber bragg grating sensors
CN114993434B (en) Vehicle weight identification system and method based on EMD energy entropy of acoustic signals

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
GR01 Patent grant
GR01 Patent grant