CN109948207A - A kind of aircraft engine high pressure rotor rigging error prediction technique - Google Patents
A kind of aircraft engine high pressure rotor rigging error prediction technique Download PDFInfo
- Publication number
- CN109948207A CN109948207A CN201910169084.4A CN201910169084A CN109948207A CN 109948207 A CN109948207 A CN 109948207A CN 201910169084 A CN201910169084 A CN 201910169084A CN 109948207 A CN109948207 A CN 109948207A
- Authority
- CN
- China
- Prior art keywords
- model
- error
- neural network
- cear
- high pressure
- 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
Abstract
The invention discloses a kind of aircraft engine high pressure rotors to assemble eccentricity prediction technique, this method is radial according to measurement mating surface and holds to runout error profile traces, feature extraction is carried out to profile traces in the method for image recognition, it establishes complex field autoregression model and obtains error locus feature vector, using seam allowance error character vector as input, eccentric position coordinate is as output, middle layer uses Gaussian function as basic function, K mean value dynamic clustering determines hidden layer center, the weight of least mean square algorithm estimation output layer constructs RBF radial basis function neural network model with this;It brings measured data into model and carries out learning training, carry out error assessment and Modifying model with finite element simulation data, determine neural network parameter, aircraft engine high pressure rotor assembly eccentricity prediction is completed with this.The present invention considers seam allowance pattern error and assembly deflections, more rapidly more accurately predicts the assembly eccentricity of high pressure rotor.
Description
Technical field
The invention belongs to mechanical assembly technique fields, and in particular to a kind of aircraft engine high pressure rotor rigging error prediction
Method.
Background technique
Aircraft engine high pressure rotor is designed as multistage-combination rotor at present.And due to the presence of part manufacturing error, turn
The rigging errors such as bias are certainly existed in son.These rigging errors will seriously affect the runnability of engine, therefore in engineering
There are extremely strict requirements to the assembly eccentricity of rotor.
Error Propagation Model is established, assembly eccentricity is predicted according to element precision testing result, assembly can be effectively reduced
The work such as trial assembly, the debugging at scene improve working efficiency.Many researchers are based on rigid body it is assumed that with homogeneous coordinate transformation
Method, Jacobian matrix method, VECTOR SCIAGRAPHY etc. carry out analysis prediction work, but these methods can not consider the pattern of mating surface
Error and assembly deflections, precision of prediction is low and each prediction is required to model again and wastes time.Also there is researcher's trial
It is predicted by the method for finite element, but finite element model is directly established by the geometrical model surveyed, number of nodes exists
Millions, scale is excessive, calculates overlong time and inconvenience.
Summary of the invention
The purpose of the present invention is to solve when the prediction of current aerospace engine high pressure rotor rigging error, rigid-body error is passed
The problems such as passing that analysis method precision of prediction is low and the Finite element arithmetic time is long and is excessively complicated proposes a kind of to know based on image
The new method of other and machine learning techniques aircraft engine high pressure rotor rigging error predictions.
In order to achieve the above objectives, the present invention is achieved by the following scheme:
A kind of aircraft engine high pressure rotor rigging error prediction technique, comprising the following steps:
Step 1: the diameter of parts and components of rotor seam allowance is jumped and the actual error data of end jump profile establish complex field autoregression,
Complex field autoregression model hereinafter referred to as CEAR model;
Step 2: passing through the feature vector of CEAR model coefficient structural configuration error locus information;
Step 3: build RBF RBF kernel function, using the feature vector of kinematic error trace information as
Input, for check offcenter amount as output, hidden layer as basic function, establishes input feature vector and output position using Gaussian function
A kind of mapping relations determine position and the width at the center of hidden layer RBF using K mean value dynamic clustering using hybrid learning scheme
Degree completes the Preliminary design of neural network with the weight of least mean square algorithm estimation output layer;
Step 4: the order of previously given CEAR model determines the dimension and hidden layer neuron of neural network input vector
Number will test the error locus data measured and the check offcenter data obtained by analysis of experimental data as input respectively
It is trained with output;
Step 5: the profile traces feature vector that the CEAR Construction of A Model of different rank on probation goes out is trained, and is recorded respectively
Time required for calculating out, COMPREHENSIVE CALCULATING efficiency primarily determine out the dimension and hidden layer neuron number of input vector;
Step 6: calculated by finite element emulation software analyze come eccentric data band enter neural network model and missed
Difference evaluation, COMPREHENSIVE CALCULATING efficiency finally determine the dimension and hidden layer neuron number of input vector, then by the data of emulation
It brings neural network into, completes the amendment to neural network;Finally, the output of neural network is exactly the eccentricity of seam allowance, completion pair
The prediction of the assembly eccentricity of parts and components of rotor.
A further improvement of the present invention lies in that establishing complex field autoregression model in step 1, the specific method is as follows:
The error information that the diameter for measuring parts and components of rotor seam allowance mating surface is jumped and end is jumped, later by profile perimeter N equal part
To N number of sampled point, sampled point P is obtainedi={ (xi, yi) | i=0,1,2 ..., N-1 } counter-profile mass center coordinate sequence, such as
Just there is plural sequence { z with complex representationi=xi+jyi| i=0,1,2 ..., N-1 }, the plural autoregression model of the ordered series of numbers indicates such as
Under
In formula:M is CEAR model order, { ak, bk, k=1,2 ... m } it is that CEAR model is
It counts and is plural number, γ is CEAR model parameter, εiFor the white noise in complex domain.
A further improvement of the present invention lies in that in step 2 tectonic movement error locus information feature vector specific method
It is as follows:
CEAR model coefficient is introduced, is enabledCan acquire the CEAR model residual error of motion profile side and
ForBy seeking local derviation dematrix equation to S, the S that must send as an envoy to is the estimated value of the smallest CEAR model coefficientThe feature vector P of CEAR model coefficient structural configuration error locus is used again.
A further improvement of the present invention lies in that constructing RBF RBF kernel function specific method such as in step 3
Under:
It builds based on RBF radial basis function neural network structure, input layer corresponds to seam allowance error locus information eigenvector
P1(a11, b11..., a1m, b1m)T、P2(a21, b21..., a2m, b2m)T, middle layer is hidden layer, using Gaussian function as base
Function determines position and the width at the center of hidden layer RBF with K mean value dynamic clustering, defeated by estimating with least mean square algorithm
The weight of layer out, output layer neuron correspond to check offcenter position coordinates, and data are respectively corresponded input and output training, complete mind
Design through network.
A further improvement of the present invention lies in that the specific method is as follows for the training of CEAR model learning in step 4:
The order of previously given CEAR model, the error locus for respectively measuring experiment carry out the data after feature extraction
It is used as to output and input with check offcenter position coordinate data and RBF radial basis function neural network is trained, provide target
Maximum frequency of training is arranged in error, records out the time required to can reach target and learning process after how many times study.
A further improvement of the present invention lies in that in step 5 RBF radial basis function neural network model learning training it is specific
Method is as follows:
The profile traces feature vector for using the CEAR Construction of A Model of different rank to go out respectively is as input to neural network mould
Type is trained, and training method builds different neural network models with step 5 respectively.
A further improvement of the present invention lies in that the specific method that RBF radial basis function neural network model is evaluated in step 6
It is as follows:
The error information that experiment is measured is as input, the check offcenter positional number calculated with finite element emulation software
Error assessment is carried out according to as output, and then to neural network model, the factors such as COMPREHENSIVE CALCULATING efficiency determine optimal nerve net
Network model order is the dimension and hidden layer neuron number for determining input vector, then using the data of emulation as input and output
It is brought into neural network, completes the amendment of neural network model.
Compared with prior art, the invention has the following advantages:
A kind of aircraft engine high pressure rotor rigging error prediction technique provided by the invention, this method is by computer graphical
Knowledge is introduced into assembly field, establishes aircraft engine high pressure rotor assembly by image recognition, the method for machine learning
Eccentricity prediction model, the eccentric prediction of completion quickly and efficiently.The letter to error profile traces is completed by image recognition
Breath extracts, and can effectively consider the relationships such as stop portion pattern error character and Planar Mechanisms cooperation, resettle artificial nerve network model
It brings experimental data and finite element simulation data into model and carries out machine learning completion modeling, assembly deflections can be incorporated and be corresponded to
In the part error of component parts, and only needs to measure profile errors during eccentric prediction later and can predict only
The eccentricity of mouth, to improve assembly efficiency.Method proposed by the present invention is higher compared with rigid-body error transmitting analysis method precision of prediction,
More complete Finite element arithmetic speed is faster.
Detailed description of the invention
Fig. 1 is to establish complex field autoregression CEAR model detailed process to the actual error of parts and components of rotor seam allowance profile
Figure.
Fig. 2 is the circular runout profile errors track of seam allowance and holds to runout error profile traces;Wherein, Fig. 2 (a) is diameter
To bounce profile errors track, Fig. 2 (b) is to hold to runout error profile traces.
Fig. 3 is the schematic diagram for designing RBF RBF kernel function.
Fig. 4 is RBF RBF kernel function training process schematic diagram.
Specific embodiment
The invention will be described in further detail with reference to the accompanying drawing:
Referring to Fig. 1, aircraft engine high pressure rotor rigging error prediction technique provided by the invention, comprising the following steps:
Step 1, it for the radial direction for the parts and components of rotor seam allowance mating surface measured and holds to error profile traces and carries out figure
As complex field autoregression model scheme is established in identification.First the run-out error of mating surface and holding to runout error is measured
The profile traces (such as Fig. 2) for obtaining the error of both direction will obtain N number of sampled point, sampled point P after profile perimeter N equal parti
={ (xi, yi) | i=0,1,2 ..., N-1 } coordinate sequence of counter-profile mass center has plural sequence { z if with complex representationi=
xi+jyi| i=0,1,2 ..., N-1 }, (circular runout profile is measured under rectangular coordinate system relative to center-of-mass coordinate, is held to jump
Driving wheel exterior feature relative to center-of-mass coordinate under cylindrical coordinates, and xiIndicate deflection angle, the y of each profile pointiIndicate each profile point
Axial runout height) the complex field autoregression model (hereinafter referred to as CEAR model) of the ordered series of numbers is represented by formula
In formula:M is CEAR model order, { ak, bk, k=1,2 ... m } it is that CEAR model is
It counts and is plural.γ is CEAR model parameter, εiFor the white noise in complex domain.
Such as profile is divided into 10 parts, presupposing CEAR model order is 5, appoints and takes γ (principle is to make εiIn [0,1] area
Between be distributed it is relatively uniform) therefore the expression of complex field autoregression model are as follows:
In formula:
Step 2, CEAR model coefficient is introduced, is enabledThe CEAR model residual error of motion profile can be acquired
Side and it isBy to S ask local derviation dematrix equation must send as an envoy to S be the smallest CEAR model coefficient estimate
Evaluation(upper official holiday set m=5), then with the feature of CEAR model coefficient structural configuration error locus to
Measure P1(a11, b11..., a1m, b1m)T、P2(a21, b21..., a2m, b2m)T
Step 3, RBF RBF kernel function is built.
Referring to fig. 2, it builds based on RBF RBF kernel function structure, as can be seen from the figure input layer pair
Answer the error locus information eigenvector P that seam allowance end is jumped and diameter is jumped1(a11, b11..., a1m, b1m)T、P2(a21, b21..., a2m,
b2m)T
Middle layer is hidden layer, using Gaussian function as basic function, the output of i-th of neural unit of hidden layer are as follows:
In formula: CiFor the central point of the Gaussian function of i-th of neural unit of hidden layer;σiIt is single for i-th of nerve of hidden layer
The width of member, position and the width at the center of hidden layer RBF are determined with K mean value dynamic clustering.(such as data center Ci,
10 samples can be selected at random from training data as data center;And for σi, since it determines radial basis function
Width, the range of observation experiment data enables radial basis function to cover entire scope, such as can be between 2-4
10 values are taken to initialize σ at randomi.Learning rate takes 0.01,0.01) change threshold of objective function takes
Output layer neuron corresponds to check offcenter position coordinates y1(Xi, Yi), y2(Xj, Yj), { (w11, w12..., whm) be
Full vector between hidden layer and output layer passes through the weight of estimation output layer with least mean square algorithm;Output layer unit it is defeated
Out are as follows:
Complete the design of neural network.
Step 4, the training of neural network model is carried out, the order of previously given CEAR model thereby determines that profile traces
Feature vector dimension, the error locus for respectively measuring experiment carry out data and check offcenter position coordinates after feature extraction
y1(Xi, Yi), y2(Xj, Yj) data as the training for carrying out model is output and input, provide target error, maximum training time is set
Number records out the time required to can reach target and learning process after how many times study.
Step 5, be trained respectively with different RBF radial basis function neural networks (such as 20 ranks are calculated separately, 30
Rank, 40 ranks, 50 ranks etc.), training method builds the neural network under different rank with step 5 respectively.
Step 6, by Xian Yuan simulation software calculate analyze come eccentric data band enter neural network progress error assessment,
The factors such as COMPREHENSIVE CALCULATING efficiency, finally determine the dimension and hidden layer neuron number of input vector, then by the data of emulation
It brings neural network into, completes the amendment to model.
Claims (7)
1. a kind of aircraft engine high pressure rotor rigging error prediction technique, which comprises the following steps:
Step 1: the diameter of parts and components of rotor seam allowance being jumped and the actual error data of end jump profile establish complex field autoregression, plural number
Domain autoregression model hereinafter referred to as CEAR model;
Step 2: passing through the feature vector of CEAR model coefficient structural configuration error locus information;
Step 3: RBF RBF kernel function is built, using the feature vector of kinematic error trace information as input,
Check offcenter amount, as basic function, establishes one kind of input feature vector and output position using Gaussian function as output, hidden layer
Mapping relations determine position and the width at the center of hidden layer RBF using K mean value dynamic clustering using hybrid learning scheme, use
Least mean square algorithm estimates the weight of output layer, completes the Preliminary design of neural network;
Step 4: the order of previously given CEAR model determines the dimension and hidden layer neuron of neural network input vector
Number, respectively will the error locus data that measure of experiment and the check offcenter data that are obtained by analysis of experimental data as input with
Output is trained;
Step 5: the profile traces feature vector that the CEAR Construction of A Model of different rank on probation goes out is trained, and records out count respectively
Time required for calculating, COMPREHENSIVE CALCULATING efficiency primarily determine out the dimension and hidden layer neuron number of input vector;
Step 6: by finite element emulation software calculate analyze come eccentric data band enter neural network model progress error comment
Valence, COMPREHENSIVE CALCULATING efficiency finally determines the dimension and hidden layer neuron number of input vector, then the data of emulation are brought into
Neural network completes the amendment to neural network;Finally, the output of neural network is exactly the eccentricity of seam allowance, is completed to rotor
The prediction of the assembly eccentricity of components.
2. aircraft engine high pressure rotor rigging error prediction technique according to claim 1, which is characterized in that step 1
It is middle to establish complex field autoregression model the specific method is as follows:
The error information that the diameter for measuring parts and components of rotor seam allowance mating surface is jumped and end is jumped will obtain N number of after profile perimeter N equal part
Sampled point obtains sampled point Pi={ (xi, yi) | i=0,1,2 ..., N-1 } counter-profile mass center coordinate sequence, such as with plural number
Expression just has plural sequence { zi=xi+jyi| i=0,1,2 ..., N-1 }, the plural autoregression model of the ordered series of numbers is expressed as follows
In formula:M is CEAR model order, { ak, bk, k=1,2 ... m } be CEAR model coefficient and
For plural number, γ is CEAR model parameter, εiFor the white noise in complex domain.
3. aircraft engine high pressure rotor rigging error prediction technique according to claim 2, which is characterized in that in step 2
The specific method is as follows for the feature vector of tectonic movement error locus information:
CEAR model coefficient is introduced, is enabledThe side of the CEAR model residual error of motion profile can be acquired and beBy seeking local derviation dematrix equation to S, the S that must send as an envoy to is the estimated value of the smallest CEAR model coefficientThe feature vector P of CEAR model coefficient structural configuration error locus is used again.
4. a kind of aircraft engine high pressure rotor rigging error prediction technique according to claim 3, which is characterized in that step
RBF RBF kernel function is constructed in rapid 3, and the specific method is as follows:
It builds based on RBF radial basis function neural network structure, input layer corresponds to seam allowance error locus information eigenvector P1
(a11, b11..., a1m, b1m)T、P2(a21, b21..., a2m, b2m)T, middle layer is hidden layer, using Gaussian function as base letter
Number, position and the width at the center of hidden layer RBF are determined with K mean value dynamic clustering, pass through estimation output with least mean square algorithm
The weight of layer, output layer neuron correspond to check offcenter position coordinates, and data are respectively corresponded input and output training, complete nerve
The design of network.
5. aircraft engine high pressure rotor rigging error prediction technique according to claim 4, which is characterized in that step 4
The specific method is as follows for middle CEAR model learning training:
The order of previously given CEAR model, the error locus for respectively measuring experiment carry out the data after feature extraction and stop
Mouth eccentric position coordinate data, which is used as to output and input, is trained RBF radial basis function neural network, provides target error,
Maximum frequency of training is set, is recorded out the time required to can reach target and learning process after how many times study.
6. aircraft engine high pressure rotor rigging error prediction technique according to claim 5, which is characterized in that step 5
The specific method is as follows for middle RBF radial basis function neural network model learning training:
Use respectively different rank CEAR Construction of A Model go out profile traces feature vector as input to neural network model into
Row training, training method build different neural network models with step 5 respectively.
7. aircraft engine high pressure rotor rigging error prediction technique according to claim 5, which is characterized in that step 6
The specific method is as follows for middle RBF radial basis function neural network model evaluation:
The error information that experiment is measured is made as input with the check offcenter position data that finite element emulation software is calculated
To export, and then error assessment is carried out to neural network model, the factors such as COMPREHENSIVE CALCULATING efficiency determine optimal neural network mould
Type order is the dimension and hidden layer neuron number for determining input vector, then the data of emulation are brought into as input and output
To neural network, the amendment of neural network model is completed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910169084.4A CN109948207A (en) | 2019-03-06 | 2019-03-06 | A kind of aircraft engine high pressure rotor rigging error prediction technique |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910169084.4A CN109948207A (en) | 2019-03-06 | 2019-03-06 | A kind of aircraft engine high pressure rotor rigging error prediction technique |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109948207A true CN109948207A (en) | 2019-06-28 |
Family
ID=67009265
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910169084.4A Pending CN109948207A (en) | 2019-03-06 | 2019-03-06 | A kind of aircraft engine high pressure rotor rigging error prediction technique |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109948207A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110530306A (en) * | 2019-08-27 | 2019-12-03 | 大连理工大学 | A kind of typical revolving parts characterizing method based on actual measurement bounce data |
CN110633521A (en) * | 2019-09-06 | 2019-12-31 | 西安交通大学 | Multistage rotor assembly phase optimization method for revolving body |
CN111178425A (en) * | 2019-12-26 | 2020-05-19 | 哈尔滨工业大学 | Surface type classification method for large-scale high-speed rotation equipment based on convolutional neural network |
CN111178506A (en) * | 2019-12-26 | 2020-05-19 | 哈尔滨工业大学 | Large-scale high-speed rotation equipment deflection and inclination eliminating method based on deep belief neural network |
CN113063589A (en) * | 2021-03-18 | 2021-07-02 | 重庆青山工业有限责任公司 | Gear microscopic error vibration prediction method based on neural network |
CN113254468A (en) * | 2021-04-20 | 2021-08-13 | 西安交通大学 | Fault query and reasoning method for certain type of equipment |
CN114440878A (en) * | 2022-01-27 | 2022-05-06 | 湖南大学无锡智能控制研究院 | SINS and DVL combined navigation method, equipment and system |
CN114782778A (en) * | 2022-04-25 | 2022-07-22 | 广东工业大学 | Assembly state monitoring method and system based on machine vision technology |
CN115223398A (en) * | 2022-07-14 | 2022-10-21 | 集美大学 | Nuclear adaptive fractional order complex value neural network AIS interpolation method considering channel constraint |
CN115630558A (en) * | 2022-12-21 | 2023-01-20 | 南京航空航天大学 | Method for predicting assembly deformation of composite material component |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5402520A (en) * | 1992-03-06 | 1995-03-28 | Schnitta; Bonnie S. | Neural network method and apparatus for retrieving signals embedded in noise and analyzing the retrieved signals |
US20020169735A1 (en) * | 2001-03-07 | 2002-11-14 | David Kil | Automatic mapping from data to preprocessing algorithms |
CN102692429A (en) * | 2011-03-24 | 2012-09-26 | 中国科学院沈阳自动化研究所 | Method for automatic identification and detection of defect in composite material |
WO2016142689A1 (en) * | 2015-03-06 | 2016-09-15 | Micromass Uk Limited | Tissue analysis by mass spectrometry or ion mobility spectrometry |
US20160314686A1 (en) * | 2013-12-30 | 2016-10-27 | Fudan University | Method for traffic flow prediction based on spatio-temporal correlation mining |
CN106741009A (en) * | 2017-01-05 | 2017-05-31 | 东南大学 | A kind of rail foreign body intrusion monitoring method based on phase sensitive optical time domain reflectometer |
CN106991212A (en) * | 2017-03-07 | 2017-07-28 | 西安交通大学 | A kind of blade root Strength Safety characteristic prediction method for optimizing GRNN network algorithms based on GA_PSO |
CN107561942A (en) * | 2017-09-12 | 2018-01-09 | 重庆邮电大学 | Intelligent vehicle track following model predictive control method based on model compensation |
CN108694294A (en) * | 2018-06-20 | 2018-10-23 | 西安交通大学 | A kind of aircraft engine high pressure rotor assembly eccentricity prediction technique |
CN108918066A (en) * | 2018-06-28 | 2018-11-30 | 东北大学 | A kind of seam allowance connection structure rotor experiment table and test method |
CN109408853A (en) * | 2018-08-31 | 2019-03-01 | 中国电力科学研究院有限公司 | A kind of power station water characteristic analysis method and system |
-
2019
- 2019-03-06 CN CN201910169084.4A patent/CN109948207A/en active Pending
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5402520A (en) * | 1992-03-06 | 1995-03-28 | Schnitta; Bonnie S. | Neural network method and apparatus for retrieving signals embedded in noise and analyzing the retrieved signals |
US20020169735A1 (en) * | 2001-03-07 | 2002-11-14 | David Kil | Automatic mapping from data to preprocessing algorithms |
CN102692429A (en) * | 2011-03-24 | 2012-09-26 | 中国科学院沈阳自动化研究所 | Method for automatic identification and detection of defect in composite material |
US20160314686A1 (en) * | 2013-12-30 | 2016-10-27 | Fudan University | Method for traffic flow prediction based on spatio-temporal correlation mining |
WO2016142689A1 (en) * | 2015-03-06 | 2016-09-15 | Micromass Uk Limited | Tissue analysis by mass spectrometry or ion mobility spectrometry |
CN106741009A (en) * | 2017-01-05 | 2017-05-31 | 东南大学 | A kind of rail foreign body intrusion monitoring method based on phase sensitive optical time domain reflectometer |
CN106991212A (en) * | 2017-03-07 | 2017-07-28 | 西安交通大学 | A kind of blade root Strength Safety characteristic prediction method for optimizing GRNN network algorithms based on GA_PSO |
CN107561942A (en) * | 2017-09-12 | 2018-01-09 | 重庆邮电大学 | Intelligent vehicle track following model predictive control method based on model compensation |
CN108694294A (en) * | 2018-06-20 | 2018-10-23 | 西安交通大学 | A kind of aircraft engine high pressure rotor assembly eccentricity prediction technique |
CN108918066A (en) * | 2018-06-28 | 2018-11-30 | 东北大学 | A kind of seam allowance connection structure rotor experiment table and test method |
CN109408853A (en) * | 2018-08-31 | 2019-03-01 | 中国电力科学研究院有限公司 | A kind of power station water characteristic analysis method and system |
Non-Patent Citations (7)
Title |
---|
T. BROTHERTON .ECT: "Automated rule extraction for engine vibration analysis", 《1999 IEEE AEROSPACE CONFERENCE. PROCEEDINGS (CAT. NO.99TH8403)》 * |
YANHUI SUN .ECT: "Repair Decision Based on Sensitivity Analysis for Aero-Engine Assembly", 《INTERNATIONAL JOURNAL OF PRECISION ENGINEERING AND MANUFACTURING》 * |
杜柳青 等: "基于图形识别的数控机床运动误差溯因方法", 《仪器仪表学报》 * |
潘立登 等: "《先进控制系统应用与维护》", 31 May 2010, 中国电力出版社 * |
熊沈蜀: "工程图矢量化处理系统", 《清华大学学报(自然科学版)》 * |
豆丁网: "《数控机床圆形轨迹误差快速查询图表讲解》", 5 April 2017 * |
顾崇衔 等: "《机械制造工艺学》", 31 August 1981, 陕西科学技术出版社 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110530306A (en) * | 2019-08-27 | 2019-12-03 | 大连理工大学 | A kind of typical revolving parts characterizing method based on actual measurement bounce data |
CN110633521A (en) * | 2019-09-06 | 2019-12-31 | 西安交通大学 | Multistage rotor assembly phase optimization method for revolving body |
CN110633521B (en) * | 2019-09-06 | 2021-07-13 | 西安交通大学 | Multistage rotor assembly phase optimization method for revolving body |
CN111178425A (en) * | 2019-12-26 | 2020-05-19 | 哈尔滨工业大学 | Surface type classification method for large-scale high-speed rotation equipment based on convolutional neural network |
CN111178506A (en) * | 2019-12-26 | 2020-05-19 | 哈尔滨工业大学 | Large-scale high-speed rotation equipment deflection and inclination eliminating method based on deep belief neural network |
CN113063589A (en) * | 2021-03-18 | 2021-07-02 | 重庆青山工业有限责任公司 | Gear microscopic error vibration prediction method based on neural network |
CN113063589B (en) * | 2021-03-18 | 2023-07-04 | 重庆青山工业有限责任公司 | Gear microscopic error vibration prediction method based on neural network |
CN113254468B (en) * | 2021-04-20 | 2023-03-31 | 西安交通大学 | Equipment fault query and reasoning method |
CN113254468A (en) * | 2021-04-20 | 2021-08-13 | 西安交通大学 | Fault query and reasoning method for certain type of equipment |
CN114440878A (en) * | 2022-01-27 | 2022-05-06 | 湖南大学无锡智能控制研究院 | SINS and DVL combined navigation method, equipment and system |
CN114440878B (en) * | 2022-01-27 | 2023-08-15 | 湖南大学无锡智能控制研究院 | SINS and DVL integrated navigation method, device and system |
CN114782778A (en) * | 2022-04-25 | 2022-07-22 | 广东工业大学 | Assembly state monitoring method and system based on machine vision technology |
CN114782778B (en) * | 2022-04-25 | 2023-01-06 | 广东工业大学 | Assembly state monitoring method and system based on machine vision technology |
CN115223398A (en) * | 2022-07-14 | 2022-10-21 | 集美大学 | Nuclear adaptive fractional order complex value neural network AIS interpolation method considering channel constraint |
CN115223398B (en) * | 2022-07-14 | 2023-09-19 | 集美大学 | Nuclear self-adaptive fractional order complex value neural network AIS interpolation method considering channel constraint |
CN115630558A (en) * | 2022-12-21 | 2023-01-20 | 南京航空航天大学 | Method for predicting assembly deformation of composite material component |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109948207A (en) | A kind of aircraft engine high pressure rotor rigging error prediction technique | |
CN108959794B (en) | Structural frequency response dynamic model correction method based on deep learning | |
Georgiou et al. | Learning fluid flows | |
CN108664690A (en) | Long-life electron device reliability lifetime estimation method under more stress based on depth belief network | |
CN108227492A (en) | A kind of discrimination method of six degree of freedom serial manipulator end load kinetic parameter | |
CN110000787A (en) | A kind of control method of super redundant mechanical arm | |
US9092737B2 (en) | Systems, methods, and apparatus for 3-D surface mapping, compliance mapping, and spatial registration with an array of cantilevered tactile hair or whisker sensors | |
CN112926152B (en) | Digital twin-driven thin-wall part clamping force precise control and optimization method | |
CN108830325A (en) | A kind of vibration information classification of landform recognition methods based on study | |
CN111638034B (en) | Strain balance temperature gradient error compensation method and system based on deep learning | |
CN109101761B (en) | Assembly contact space attitude determination method based on actually measured surface morphology | |
Rao et al. | Normal direction measurement and optimization with a dense three-dimensional point cloud in robotic drilling | |
CN114169383A (en) | Strain-displacement reconstruction method of finite element model structure | |
CN103823406A (en) | Numerical control machine tool sensitive-link identification method based on modal mass distribution matrix | |
CN108846200B (en) | Quasi-static bridge influence line identification method based on iteration method | |
CN114218778A (en) | Method and device for analyzing sonic boom test data | |
CN105354850A (en) | Method for detecting size of complex curved part in three-dimensional matching manner based on electric-field properties | |
CN108920757A (en) | A kind of Problem in Vehicle Crash Accident Reconstruction method based on Computer Simulation | |
CN106354935B (en) | Complex curved surface parts matching detection method based on electron outside nucleus probability density distribution | |
Li et al. | A remaining useful life prediction method considering the dimension optimization and the iterative speed | |
CN110110406B (en) | Slope stability prediction method for achieving LS-SVM model based on Excel computing platform | |
US11958112B2 (en) | Characterizing liquid reflective surfaces in 3D liquid metal printing | |
US20220414380A1 (en) | Characterizing liquid reflective surfaces in 3d liquid metal printing | |
CN114638048A (en) | Three-dimensional spray pipe flow field rapid prediction and sensitivity parameter analysis method and device | |
Adams et al. | Validation principles of agricultural machine multibody dynamics models |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190628 |
|
RJ01 | Rejection of invention patent application after publication |