CN102034111A - Method for identifying and detecting aircraft structural damage conditions in diversified way - Google Patents
Method for identifying and detecting aircraft structural damage conditions in diversified way Download PDFInfo
- Publication number
- CN102034111A CN102034111A CN 201010589193 CN201010589193A CN102034111A CN 102034111 A CN102034111 A CN 102034111A CN 201010589193 CN201010589193 CN 201010589193 CN 201010589193 A CN201010589193 A CN 201010589193A CN 102034111 A CN102034111 A CN 102034111A
- Authority
- CN
- China
- Prior art keywords
- energy
- wavelet
- wavelet packet
- signal
- degree
- 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
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a method for identifying and detecting aircraft structural damage conditions in a diversified way and relates to a multi-method combined system for identifying and detecting the aircraft structural damage conditions, which is characterized by being based on vibration test signals of an aircraft structure, applying a wavelet packet method and being used for identifying spatial position of a structural damage and damage degree; meanwhile, a wavelet packet energy gradient index is provided, accurate quantitative analysis is carried out on the damage degree of the structure by combining a neural network and taking the wavelet packet energy gradient index as network input; and moreover, a grey prediction method is used for wavelet packet energy index prediction to further carry out quantitative prediction analysis on the structural damage conditions. The method has the advantages of rapidness, high precision, implementation predictability and functional diversity, is beneficial to the performing of the combined advantages of the high-frequency resolution of the wavelet packet, the nonlinear mapping of the BP (Back Propagation) neural network and the predictability of the grey prediction, and improves the capability of identifying and predicting the aircraft structural damage.
Description
Technical field
The present invention relates to a kind of method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects.
Background technology
The variation of structural parameters is embodied directly in the change of physical parameter, causes the variation of characteristic parameters such as its natural frequency, the vibration shape, and then the dynamic response of structure is changed.Directly, can avoid that structure is carried out numerous and diverse modal parameter and find the solution based on vibration monitoring of engineering structure signal recognition structure parameter situation of change.These class methods adopt signal processing methods such as Fourier transform, wavelet transformation directly the structural response signal to be analyzed and handled, according to the variation or the statistical property of signal parameter, and the parameter situation of change of recognition structure.Compare with Fourier transform, wavelet analysis has multiple dimensioned characteristic, all has the ability of characterization signal local feature in time domain and frequency domain, can detect structural parameters and change the feeble signal sudden change that causes.
Artificial neural network be to human brain abstract, simplify and simulation, the fundamental characteristics of reflection human brain can be finished functions such as study, memory, identification and reasoning.From systematic point of view, artificial neural network is the self-adaptation nonlinear dynamic system that is formed by extremely abundant and perfect connection by a large amount of neurons.Connected mode difference between the neuron, the structural form of neural network are also just different.Artificial neural network has following feature: (1) distributes and stores and fault-tolerance; (2) massively parallel processing; (3) self study, self-organization and adaptivity; (4) show the feature of complicated nonlinear dynamic system; (5) can handle the unclear and indefinite problem of inference rule of some environmental information complexity, knowledge background.
Consider the energy parameter change indicator of some structural damage, face that sample is few, uncertain big, determinacy analytical Prediction model is difficult to difficult problems such as foundation, this is just meeting the characteristics that gray system theory studies a question, and promptly sequentiality, minority are according to property, perfect information, time transitivity, grey causality.
After parameter variation (damage) appears in Flight Vehicle Structure, corresponding the changing of response of structure signal meeting, damage changes to the inhibition and the humidification of each radio-frequency component.Usually, it can play inhibiting effect to some radio-frequency component significantly, and the other radio-frequency component is played humidification.Therefore, the response output during structural damage is compared with structure response output just often, and the energy of signal has bigger difference in the same frequency band, and signal energy reduces in some frequency band, and the signal energy of some other frequency band increases.So, in the energy of each radio-frequency component signal, comprising abundant structures parameter change information, a kind of degree of impairment has promptly been represented in the change of certain or certain several radio-frequency component energy.
Summary of the invention
Since the aerocraft system complexity, and structural parameters variation kind is also varied, and the data volume of processing is bigger.Accurately quantize for the parameter to Flight Vehicle Structure changes degree of injury, obtain characterizing the proper vector of structural damage, the present invention proposes a kind of method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects.This method is based on the vibration-testing signal of Flight Vehicle Structure, use method of wavelet packet, be used for locus and degree of injury that the recognition structure damage takes place, a kind of wavelet-packet energy rate of change index is proposed simultaneously, in conjunction with neural network, with the wavelet-packet energy index is that the network input is carried out accurately quantitative analysis to damage of structure, also adopts the gray theory model method to carry out the wavelet-packet energy index prediction, carries out the structural damage state and quantizes forecast analysis.The present invention has rapidity, high precision and realizes the advantage of predictability, functional diversity, help bringing into play the high frequency resolution of wavelet packet, the Nonlinear Mapping of BP neural network and the advantage that gray theory model prediction ability combines, improve the identification and the predictive ability of Flight Vehicle Structure damage.
A kind of method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects of the present invention is characterized in that this method comprises the following steps:
The first step: to the displacement structure response signal of Flight Vehicle Structure output
Handle by WAVELET PACKET DECOMPOSITION, obtain the energy component of structural vibration displacement response signal
(abbreviate the signal energy component as
);
Second step: adopt the energy gradient index model
A is an original state, and b is the state under the current time, to the signal energy component
Handle, obtain wavelet energy rate of change index WEI
F (t) j(abbreviate small echo-rate of change WEI as
F (t) j), described small echo-rate of change WEI
F (t) jCan be used in the locus A that identification Flight Vehicle Structure damage takes place (x, y, z);
The 3rd step: adopt the BP neural network algorithm to small echo-rate of change WEI
F (t) jHandle, obtain wavelet energy-degree of injury WDD
F (t), this wavelet energy-degree of injury WDD
F (t)Wavelet energy-the degree of injury under the current time just;
The 4th step: adopt gray system theory to the signal energy component
Handle, obtain next signal energy component constantly
The 5th step: adopt the energy gradient index model
To next signal energy component constantly
Handle, obtain next small echo-rate of change WEI constantly
F (t+1) j
The 6th step: adopt the BP neural network algorithm to next small echo-rate of change WEI constantly
F (t+1) jHandle, obtain next wavelet energy-degree of injury NWDD constantly
F (t), this next wavelet energy-degree of injury NWDD constantly
F (t)Wavelet energy-the degree of injury that may occur just.
Diversification identification of the present invention is with the advantage of the method that detects:
1. overcome the deficiency of existing monotechnics, provide a kind of based on vibration monitoring of engineering structure signal recognition structure degree of injury efficiently, the system that can carry out the degree of injury prediction simultaneously.Utilize the wavelet-packet energy index, the damage status of accurate description scheme, utilize the non-linear approximation capability of neural network to set up mapping relations between faulted condition index and the degree of injury, the sequentiality, minority that utilize the gray theory model are predicted energy parameter change indicator data according to advantages such as property, perfect information, time transitivity, grey causalities.This method synthesis utilizes the advantage of the whole bag of tricks, help bringing into play the high frequency resolution of wavelet packet, the Nonlinear Mapping of BP neural network and the advantage that gray theory model prediction ability combines, utilize the wavelet packet gross energy rate of change index that proposes, carry out the quantitative analysis of structural damage degree fast, utilize the gray theory model that wavelet packet gross energy rate of change achievement data is predicted, the forecast function of implementation structure damage.
2. it is directly perceived, easy to utilize wavelet transformation to carry out parameter variation identification, thereby obtains paying attention in the online health monitoring of Flight Vehicle Structure.
3. the present invention adopts " energy-damage " to carry out the identification of structural parameters changing pattern, this method does not need structure is carried out modeling, and the direct independent reconstruct of Flight Vehicle Structure sequential response signal being carried out WAVELET PACKET DECOMPOSITION and each constituent signals, therefrom extract the energy parameter change indicator then, determine position and degree of injury that structural parameters change.
Description of drawings
Fig. 1 is identification of Flight Vehicle Structure faulted condition and the process flow diagram that detects.
Fig. 2 is a BP neural metwork training synoptic diagram of the present invention.
Embodiment
The present invention is described in further detail below in conjunction with accompanying drawing.
A kind of method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects of the present invention, this method comprises the following steps:
The first step: to the displacement structure response signal of Flight Vehicle Structure output
Handle by WAVELET PACKET DECOMPOSITION, obtain the energy component of structural vibration displacement response signal
(abbreviate the signal energy component as
);
Second step: adopt the energy gradient index model
To the signal energy component
Handle, obtain wavelet energy rate of change index WEI
F (t) j(abbreviate small echo-rate of change WEI as
F (t) j), described small echo-rate of change WEI
F (t) jCan be used in the locus A that identification Flight Vehicle Structure damage takes place (x, y, z);
In the present invention, the signal gross energy can be regarded the energy sum of each component of wavelet packet in the different frequency bands as.Owing to comprise the pollution that the small echo component of less energy is subjected to measuring noise easily, in analytic process, will give up to them.Energy to each component after the WAVELET PACKET DECOMPOSITION carries out obtaining the energy gradient index model from big extremely little ordering
A is an original state, and b is the state under the current time, and the energy gradient before and after changing by the structure dynamic performance parameters shows the health status of structure.A, the corresponding two states of b, this index has reflected the relative intensity of variation of structural physical parameter variation (damage) front and back structural response signal wavelet energies.It not only changes relevant, also relevant with load-up condition with the structure dynamic performance parameters.So, change the stand under load situation unanimity that will guarantee the structural damage front and back in the identifying at the structure dynamic performance parameters.
The 3rd step: adopt the BP neural network algorithm to small echo-rate of change WEI
F (t) jHandle, obtain wavelet energy-degree of injury WDD
F (t), this wavelet energy-degree of injury WDD
F (t)Wavelet energy-the degree of injury under the current time just;
The 4th step: adopt gray system theory to the signal energy component
Handle, obtain next signal energy component constantly
In the present invention, adopt the gray theory model that the different wavelet-packet energy achievement datas constantly of Flight Vehicle Structure displacement response signal extraction point are predicted, the wavelet-packet energy index that obtains is imported existing BP neural network carry out forecast analysis.Existing wavelet-packet energy rate of change achievement data is utilized the gray theory models treated, obtain next wavelet-packet energy rate of change index constantly, network with next index input constantly trains obtains next lesion quantification state constantly, has promptly carried out the lesion quantification prediction.
The 5th step: adopt the energy gradient index model
To next signal energy component constantly
Handle, obtain next small echo-rate of change WEI constantly
F (t+1) j
The 6th step: adopt the BP neural network algorithm to next small echo-rate of change WEI constantly
F (t+1) jHandle, obtain next wavelet energy-degree of injury NWDD constantly
F (t), this next wavelet energy-degree of injury NWDD constantly
F (t)Wavelet energy-the degree of injury that may occur just.
In the present invention, the flow process of whole Flight Vehicle Structure faulted condition identification and detection as shown in Figure 1.Based on the vibration-testing of Flight Vehicle Structure, obtain the displacement structure response signal
Right
Carry out WAVELET PACKET DECOMPOSITION, the picked up signal energy component
Utilize the signal energy component
Adopt the energy gradient index model to obtain small echo-rate of change WEI
F (t) j, with the small echo-rate of change WEI that obtains
F (t) jCan carry out wavelet energy-degree of injury WDD in conjunction with the BP neural network algorithm
F (t)Identification, adopt the signal energy component of gray system theory model simultaneously to current time
Handle and obtain next signal energy component constantly
Adopt the energy gradient index model to obtain small echo-rate of change WEI again
F (t+1) j, with next the small echo-rate of change WEI constantly that obtains
F (t+1) jAdopt the BP neural network algorithm can obtain next wavelet energy-degree of injury NWDD of Flight Vehicle Structure constantly
F (t)
In the first step of the present invention, the energy component detailed process that described first step WAVELET PACKET DECOMPOSITION is extracted the structural response signal is:
Step 1-1: to the displacement response signal that vibration monitoring of engineering structure obtained
Carry out j layer WAVELET PACKET DECOMPOSITION, obtain displacement response signal after the WAVELET PACKET DECOMPOSITION
T is vibration signal time test period, and j is the WAVELET PACKET DECOMPOSITION number of plies, and i is the sequence number at j layer medium frequency layer, and i=1,2 ... 2
j
Wherein
Be wavelet packet basis functions
Linear combination k=(∞ ,+∞), promptly
Be wavelet packet coefficient;
In the present invention, wavelet packet coefficient
Depend on signal
With wavelet basis function
Registration, then have
Step 1-2: calculate the wavelet packet signal energy under each decomposition level
Under j layer WAVELET PACKET DECOMPOSITION, signal
Energy be designated as
According to the orthogonality condition of wavelet basis function, have gross energy and equal each component energy sum
M is the frequency layer identification number of a preceding iteration, m ∈ i,
Decomposed signal for a preceding iteration frequency layer under j layer WAVELET PACKET DECOMPOSITION;
N is the frequency layer identification number of last time iteration, n ∈ i,
Decomposed signal for previous iteration frequency layer under j layer WAVELET PACKET DECOMPOSITION;
Dt is the differential of time;
In the present invention, described BP neural metwork training step is:
Step 3-1: the number of plies of network
Having deviation and at least one S type hidden layer adds and can approach the network of a linear output layer any rational function, thereby construct 3 layers of neural network.
Step 3-2: the neuron number of hidden layer
The training precision of network improves by the neuron number that increases in the hidden layer, selects according to practical application.In the present invention, the neuron number is set to 17.
Step 3-3: the choosing of initial weight
Because nerve network system is non-linear, initial value for e-learning whether reach local minimum, whether can restrain and the length relation of training time very big.If initial value is too big, make that the input after the weighting drops on the saturation region of activation function, thereby cause the weighting correction factor very little that the network adjustment process almost is deadlocked.In the present invention, initial weight is generally got the random number in (1,1).
Step 3-4: learning rate
Learning rate has determined the weights variable quantity of network training generation each time.Learning rate is crossed the instability that conference causes system, but learning rate is too little, and the training time of network is longer, and speed of convergence is very slow, but can guarantee that the error of network can finally be tending towards minimum error values.Therefore, tend to choose less learning rate generally speaking to guarantee the stability of system, the scope of choosing of learning rate of the present invention is between 0.01～0.8.
Step 3-5: the choosing of anticipation error
In the training process of network, anticipation error should be set at a suitable value.So-called " suitable " is to determine with respect to the node number of needed hidden layer.Because less expected error value requires to increase the number and the net training time of hidden layer node.Usually as a comparison, can train the network of two different expected error value simultaneously, at last by taking all factors into consideration to determine to adopt one of them network.
In the present invention, utilize small echo-rate of change WEI
F (t) jAchievement data, training BP neural network is obtained the node weights of network, process as shown in Figure 2.In conjunction with the BP neural network, be the network input with the wavelet-packet energy index, damage of structure is carried out accurately quantitative analysis.Adopt three layers of BP neural network that the degree of injury of Flight Vehicle Structure is carried out quantitative research.The wavelet-packet energy rate of change index that is input as structural response of neural network, export structure degree of injury index WDD
F (t), be nondimensional number percent.
The present invention adopts the diversification method to overcome the deficiency of existing monotechnics, provides a kind of based on vibration monitoring of engineering structure signal recognition structure degree of injury efficiently, the system that can carry out the degree of injury prediction simultaneously.Utilize the wavelet-packet energy index, the damage status of accurate description scheme, utilize the non-linear approximation capability of neural network to set up mapping relations between faulted condition index and the degree of injury, the sequentiality, minority that utilize the gray theory model are predicted energy parameter change indicator data according to advantages such as property, perfect information, time transitivity, grey causalities.This method synthesis utilizes the advantage of the whole bag of tricks, help bringing into play the high frequency resolution of wavelet packet, the Nonlinear Mapping of BP neural network and the advantage that gray theory model prediction ability combines, utilize the wavelet packet gross energy rate of change index that proposes, carry out the quantitative analysis of structural damage degree fast, utilize the gray theory model that wavelet packet gross energy rate of change achievement data is predicted, the forecast function of implementation structure damage.
Claims (3)
1. the method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects is characterized in that this method comprises the following steps:
The first step: to the displacement structure response signal of Flight Vehicle Structure output
Handle by WAVELET PACKET DECOMPOSITION, obtain the signal energy component
Second step: adopt the energy gradient index model
A is an original state, and b is the state under the current time, to the signal energy component
Handle, obtain small echo-rate of change WEI
F (t) j
Described small echo-rate of change WEI
F (t) jCan be used in the locus A that identification Flight Vehicle Structure damage takes place (x, y, z);
The 3rd step: adopt the BP neural network algorithm to small echo-rate of change WEI
F (t) jHandle, obtain wavelet energy-degree of injury WDD
F (t), this wavelet energy-degree of injury WDD
F (t)Wavelet energy-the degree of injury under the current time just;
The 4th step: adopt gray system theory to the signal energy component
Handle, obtain next signal energy component constantly
The 5th step: adopt the energy gradient index model
To next signal energy component constantly
Handle, obtain next small echo-rate of change WEI constantly
F (t+1) j
The 6th step: adopt the BP neural network algorithm to next small echo-rate of change WEI constantly
F (t+1) jHandle, obtain next wavelet energy-degree of injury NWDD constantly
F (t), this next wavelet energy-degree of injury NWDD constantly
F (t)Wavelet energy-the degree of injury that may occur just.
2. the method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects according to claim 1, it is characterized in that: the energy component detailed process that described first step WAVELET PACKET DECOMPOSITION is extracted the structural response signal is:
Step 1-1: to the displacement response signal that vibration monitoring of engineering structure obtained
Carry out j layer WAVELET PACKET DECOMPOSITION, obtain displacement response signal after the WAVELET PACKET DECOMPOSITION
T is vibration signal time test period, and j is the WAVELET PACKET DECOMPOSITION number of plies, and i is the sequence number at j layer medium frequency layer, and i=1,2 ... 2j;
Wherein
Be wavelet packet basis functions
Linear combination k=(∞ ,+∞), promptly
Be wavelet packet coefficient;
Described wavelet packet coefficient
Depend on signal
With wavelet basis function
Registration, then have
Step 1-2: calculate the wavelet packet signal energy under each decomposition level
Under j layer WAVELET PACKET DECOMPOSITION, signal
Energy be designated as
According to the orthogonality condition of wavelet basis function, have gross energy and equal each component energy sum
M is the frequency layer identification number of a preceding iteration, m ∈ i,
Decomposed signal for a preceding iteration frequency layer under j layer WAVELET PACKET DECOMPOSITION;
N is the frequency layer identification number of last time iteration, n ∈ i,
Decomposed signal for previous iteration frequency layer under j layer WAVELET PACKET DECOMPOSITION;
Dt is the differential of time;
3. the method that the Flight Vehicle Structure faulted condition is carried out diversification identification and detects according to claim 1, it is characterized in that: the BP neural network is 3 layers of neural network, and hidden layer neuron node number is 17.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010589193 CN102034111A (en) | 2010-12-15 | 2010-12-15 | Method for identifying and detecting aircraft structural damage conditions in diversified way |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010589193 CN102034111A (en) | 2010-12-15 | 2010-12-15 | Method for identifying and detecting aircraft structural damage conditions in diversified way |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102034111A true CN102034111A (en) | 2011-04-27 |
Family
ID=43886974
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010589193 Pending CN102034111A (en) | 2010-12-15 | 2010-12-15 | Method for identifying and detecting aircraft structural damage conditions in diversified way |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102034111A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104458173A (en) * | 2014-11-27 | 2015-03-25 | 广东电网有限责任公司中山供电局 | Steel framework structure mutational damage recognition method and system |
| CN105225223A (en) * | 2015-08-27 | 2016-01-06 | 南京市计量监督检测院 | Based on the damage Detection of Smart Composite Structure method of wavelet analysis and BP neural network |
| CN108613737A (en) * | 2018-05-14 | 2018-10-02 | 南京理工大学 | The discrimination method of aircraft multifrequency vibration signal based on wavelet packet and STFT |
| CN109815940A (en) * | 2019-03-05 | 2019-05-28 | 韦灼彬 | Wavelet-packet energy spectrometry damnification recognition method |
| CN110908365A (en) * | 2019-12-25 | 2020-03-24 | 西北工业大学 | Unmanned aerial vehicle sensor fault diagnosis method and system and readable storage medium |
| CN111307944A (en) * | 2020-03-15 | 2020-06-19 | 中国飞机强度研究所 | Quantitative monitoring method and system for structural damage of composite material |
| CN111581865A (en) * | 2020-05-08 | 2020-08-25 | 成都山地环安防灾减灾技术有限公司 | Remote monitoring and early warning method and system for engineering structure damage |
| CN113450333A (en) * | 2021-06-30 | 2021-09-28 | 哈尔滨工业大学 | Machine learning-based reinforced concrete column earthquake damage degree evaluation method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101063643A (en) * | 2007-02-02 | 2007-10-31 | 北京航空航天大学 | Intelligent diagnostic method for airplane functional failure and system thereof |
| CN101201295A (en) * | 2006-12-13 | 2008-06-18 | 上海海事大学 | Method and device for predicting grey failure of rotating machinery wavelet |
-
2010
- 2010-12-15 CN CN 201010589193 patent/CN102034111A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101201295A (en) * | 2006-12-13 | 2008-06-18 | 上海海事大学 | Method and device for predicting grey failure of rotating machinery wavelet |
| CN101063643A (en) * | 2007-02-02 | 2007-10-31 | 北京航空航天大学 | Intelligent diagnostic method for airplane functional failure and system thereof |
Non-Patent Citations (4)
| Title |
|---|
| 《中国优秀硕士学位论文全文数据库》 20091031 陈长万 基于小波包能量和神经网络的结构损伤识别法研究 1-3 , * |
| 《北京力学会第十六届学术年会论文集》 20100110 孙伶俐 等 基于BP神经网络的飞行器结构损伤量化分析 434-435 , * |
| 《地震工程与工程振动》 20031031 李宏男 等 基于小波分析和神经网络的框架结构损伤诊断方法 141-148 第23卷, 第5期 * |
| 《第十二届全国设备故障诊断学术会议论文集》 20100815 孙伶俐 等 基于遥测数据的飞行器故障预测若干方法对比 53-55 , * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104458173A (en) * | 2014-11-27 | 2015-03-25 | 广东电网有限责任公司中山供电局 | Steel framework structure mutational damage recognition method and system |
| CN105225223A (en) * | 2015-08-27 | 2016-01-06 | 南京市计量监督检测院 | Based on the damage Detection of Smart Composite Structure method of wavelet analysis and BP neural network |
| CN108613737A (en) * | 2018-05-14 | 2018-10-02 | 南京理工大学 | The discrimination method of aircraft multifrequency vibration signal based on wavelet packet and STFT |
| CN109815940A (en) * | 2019-03-05 | 2019-05-28 | 韦灼彬 | Wavelet-packet energy spectrometry damnification recognition method |
| CN110908365A (en) * | 2019-12-25 | 2020-03-24 | 西北工业大学 | Unmanned aerial vehicle sensor fault diagnosis method and system and readable storage medium |
| CN111307944A (en) * | 2020-03-15 | 2020-06-19 | 中国飞机强度研究所 | Quantitative monitoring method and system for structural damage of composite material |
| CN111581865A (en) * | 2020-05-08 | 2020-08-25 | 成都山地环安防灾减灾技术有限公司 | Remote monitoring and early warning method and system for engineering structure damage |
| CN111581865B (en) * | 2020-05-08 | 2023-09-05 | 成都山地环安科技有限公司 | Engineering structure damage remote monitoring and early warning method and system |
| CN113450333A (en) * | 2021-06-30 | 2021-09-28 | 哈尔滨工业大学 | Machine learning-based reinforced concrete column earthquake damage degree evaluation method |
| CN113450333B (en) * | 2021-06-30 | 2022-01-28 | 哈尔滨工业大学 | Machine learning-based reinforced concrete column earthquake damage degree evaluation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102034111A (en) | Method for identifying and detecting aircraft structural damage conditions in diversified way | |
| CN108520199B (en) | Human body action open set identification method based on radar image and generation countermeasure model | |
| CN106779151B (en) | A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method | |
| CN112685961B (en) | Method and system for predicting remaining service life of analog circuit | |
| CN103728551B (en) | A kind of analog-circuit fault diagnosis method based on cascade integrated classifier | |
| CN109389058A (en) | Sea clutter and noise signal classification method and system | |
| CN108090295A (en) | A kind of long-span cablestayed bridges Damages in Stay Cables recognition methods | |
| CN110070229A (en) | The short term prediction method of home electrical load | |
| CN109034054B (en) | Harmonic multi-label classification method based on LSTM | |
| CN103926526A (en) | Analog circuit fault diagnosis method based on improved RBF neural network | |
| CN110161370A (en) | A kind of electric network fault detection method based on deep learning | |
| CN104459668A (en) | Radar target recognition method based on deep learning network | |
| CN103268525B (en) | A kind of Hydrological Time Series simulating and predicting method based on WD-RBF | |
| CN103995237A (en) | Satellite power supply system online fault diagnosis method | |
| CN102520342A (en) | Analog circuit test node selecting method based on dynamic feedback neural network modeling | |
| CN102034133A (en) | Quantum neural network-based comprehensive evaluation method for multi-factor system | |
| CN109214356A (en) | A kind of fan transmission system intelligent fault diagnosis method based on DCNN model | |
| CN107993012A (en) | A kind of adaptive electric system on-line transient stability appraisal procedure of time | |
| CN110298374A (en) | A kind of driving locus energy consumption analysis method and apparatus based on deep learning | |
| CN106656357A (en) | System and method of evaluating state of power frequency communication channel | |
| CN109142976A (en) | Cable fault examination method and device | |
| CN114781435A (en) | Power electronic circuit fault diagnosis method based on improved Harris eagle optimization algorithm optimized variation modal decomposition | |
| CN113392773A (en) | Transformer fault diagnosis system of convolutional neural network based on map information fusion | |
| CN104933471A (en) | Phase space reconstruction method for use in short-term wind speed prediction method | |
| CN110825583B (en) | Energy efficiency qualitative assessment technology for multi-index fusion of cloud data center |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110427 |