Transformer winding modal parameter identification based on poly-reference least-square complex frequency domain method

Similar documents
Time-frequency Analysis Based State Diagnosis of Transformers Windings under the Short-Circuit Shock

To: Professor Avitabile Date: February 4, 2003 From: Mechanical Student Subject: Experiment #1 Numerical Methods Using Excel

Sensors for Motion and Position Measurement

Low Sampling Rate Technology for UHF Partial Discharge Signals Based on Sparse Vector Recovery

Uncertainty in measurements of power and energy on power networks

Control of Chaos in Positive Output Luo Converter by means of Time Delay Feedback

29. Network Functions for Circuits Containing Op Amps

antenna antenna (4.139)

Research of Dispatching Method in Elevator Group Control System Based on Fuzzy Neural Network. Yufeng Dai a, Yun Du b

A High-Sensitivity Oversampling Digital Signal Detection Technique for CMOS Image Sensors Using Non-destructive Intermediate High-Speed Readout Mode

Research on Peak-detection Algorithm for High-precision Demodulation System of Fiber Bragg Grating

Chaotic Filter Bank for Computer Cryptography

RC Filters TEP Related Topics Principle Equipment

ANNUAL OF NAVIGATION 11/2006

Research on Controller of Micro-hydro Power System Nan XIE 1,a, Dezhi QI 2,b,Weimin CHEN 2,c, Wei WANG 2,d

Graph Method for Solving Switched Capacitors Circuits

Optimization Frequency Design of Eddy Current Testing

Introduction. (Received 08 January 2009; accepted 10 March 2009)

NATIONAL RADIO ASTRONOMY OBSERVATORY Green Bank, West Virginia SPECTRAL PROCESSOR MEMO NO. 25. MEMORANDUM February 13, 1985

Numerical simulations for long range guided waves Nondestructive Testing by a wavelet based two port equivalent.

Rejection of PSK Interference in DS-SS/PSK System Using Adaptive Transversal Filter with Conditional Response Recalculation

A Proposal of Mode Shape Estimation Method Using Pseudo-Modal Response : Applied to Steel Bridge in Building

Active and Reactive Power Control of DFIG for Wind Energy Conversion Using Back to Back Converters (PWM Technique)

Implementation of Fan6982 Single Phase Apfc with Analog Controller

Figure.1. Basic model of an impedance source converter JCHPS Special Issue 12: August Page 13

Pneumatic-Piezoelectric Hybrid Vibration Suppression For a Flexible Translating Beam Using Adaptive Fuzzy Sliding Mode Control Algorithm

FPGA Implementation of Ultrasonic S-Scan Coordinate Conversion Based on Radix-4 CORDIC Algorithm

Application of Intelligent Voltage Control System to Korean Power Systems

IEE Electronics Letters, vol 34, no 17, August 1998, pp ESTIMATING STARTING POINT OF CONDUCTION OF CMOS GATES

MTBF PREDICTION REPORT

TECHNICAL NOTE TERMINATION FOR POINT- TO-POINT SYSTEMS TN TERMINATON FOR POINT-TO-POINT SYSTEMS. Zo = L C. ω - angular frequency = 2πf

Calculation of the received voltage due to the radiation from multiple co-frequency sources

Equivalent Circuit Model of Electromagnetic Behaviour of Wire Objects by the Matrix Pencil Method

Th P5 13 Elastic Envelope Inversion SUMMARY. J.R. Luo* (Xi'an Jiaotong University), R.S. Wu (UC Santa Cruz) & J.H. Gao (Xi'an Jiaotong University)

Fault Detection of Aircraft Cable via Spread Spectrum Time Domain Reflectometry

Estimation of Synchronous Generator Parameters from On-line Measurements

PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION. Evgeny Artyomov and Orly Yadid-Pecht

Performance Analysis of Multi User MIMO System with Block-Diagonalization Precoding Scheme

Discrete Time Sliding Mode Control of Magnetic Levitation System with Enhanced Exponential Reaching Law

INSTANTANEOUS TORQUE CONTROL OF MICROSTEPPING BIPOLAR PWM DRIVE OF TWO-PHASE STEPPING MOTOR

A study of turbo codes for multilevel modulations in Gaussian and mobile channels

CMOS Implementation of Lossy Integrator using Current Mirrors Rishu Jain 1, Manveen Singh Chadha 2 1, 2

Design of Teaching Platform Based on Information Detection System

Development of a High Bandwidth, High Power Linear Amplifier for a Precision Fast Tool Servo System

MODAL PARAMETER ESTIMATION FOR OPERATIONAL WIND TURBINES

Harmonic Modeling of Inrush Current in Core Type Power Transformers using Hartley Transform

Two-Phase Asynchronous Motor - Simulation and Measurement

Romuald Rzadkowski. Vitally Gnesin, Luba Kolodyazhnaya. Ryszard Szczepanik

Walsh Function Based Synthesis Method of PWM Pattern for Full-Bridge Inverter

A Current Differential Line Protection Using a Synchronous Reference Frame Approach

Estimating Mean Time to Failure in Digital Systems Using Manufacturing Defective Part Level

Boise State University Department of Electrical and Computer Engineering ECE 212L Circuit Analysis and Design Lab

POLYTECHNIC UNIVERSITY Electrical Engineering Department. EE SOPHOMORE LABORATORY Experiment 1 Laboratory Energy Sources

A method to reduce DC-link voltage fluctuation of PMSM drive system with reduced DC-link capacitor

Error of Varhour Meter s Registration in the Presence of Harmonics

Open Access Research on PID Controller in Active Magnetic Levitation Based on Particle Swarm Optimization Algorithm

STUDY OF MATRIX CONVERTER BASED UNIFIED POWER FLOW CONTROLLER APPLIED PI-D CONTROLLER

Fault Detection and Diagnosis in Induction Machines: A Case Study

Evaluate the Effective of Annular Aperture on the OTF for Fractal Optical Modulator

High Speed, Low Power And Area Efficient Carry-Select Adder

A Shunt Active Power Filter with Enhanced Dynamic Performance using Dual-Repetitive Controller and Predictive Compensation

Research on control algorithm of high voltage circuit breaker three phase permanent magnet brushless DC motor

FAST ELECTRON IRRADIATION EFFECTS ON MOS TRANSISTOR MICROSCOPIC PARAMETERS EXPERIMENTAL DATA AND THEORETICAL MODELS

Advances in Structural Engineering

The Application of Interpolation Algorithms in OFDM Channel Estimation

Digital Differential Protection of Power Transformer Using Matlab

ScienceDirect. Vibration Control of a Cable-Stayed Footbridge Using the Tension Changes of Cable

THE ARCHITECTURE OF THE BROADBAND AMPLIFIERS WITHOUT CLASSICAL STAGES WITH A COMMON BASE AND A COMMON EMITTER

A New Type of Weighted DV-Hop Algorithm Based on Correction Factor in WSNs

Microelectronic Circuits

DETERMINATION OF WIND SPEED PROFILE PARAMETERS IN THE SURFACE LAYER USING A MINI-SODAR

Reconstruction of the roadway coverage parameters from radar probing measurements

Fast Code Detection Using High Speed Time Delay Neural Networks

A Comparison of Two Equivalent Real Formulations for Complex-Valued Linear Systems Part 2: Results

Small Range High Precision Positioning Algorithm Based on Improved Sinc Interpolation

A Novel Acoustic Emission Source Location Method in the Crane Based on EEMD-FastICA

THE GENERATION OF 400 MW RF PULSES AT X-BAND USING RESONANT DELAY LINES *

An experimental interface design for a single-link elastic manipulator system

Comparative Analysis of Reuse 1 and 3 in Cellular Network Based On SIR Distribution and Rate

Optimal Placement of PMU and RTU by Hybrid Genetic Algorithm and Simulated Annealing for Multiarea Power System State Estimation

Micro-grid Inverter Parallel Droop Control Method for Improving Dynamic Properties and the Effect of Power Sharing

A Preliminary Study on Targets Association Algorithm of Radar and AIS Using BP Neural Network

An Adaptive Over-current Protection Scheme for MV Distribution Networks Including DG

Shunt Active Filters (SAF)

Latency Insertion Method (LIM) for IR Drop Analysis in Power Grid

Digital Transmission

[Type text] [Type text] [Type text] Wenjing Yuan Luxun Art Academy of Yan an University Xi an, , (CHINA)

Strain Gauge Measuring Amplifier BA 660

Model Reference Current Control of a Unipolar Induction Motor Drive

Performance Analysis of Power Line Communication Using DS-CDMA Technique with Adaptive Laguerre Filters

Modelling and performance analysis of dual-channel switched reluctance motor

Passive Filters. References: Barbow (pp ), Hayes & Horowitz (pp 32-60), Rizzoni (Chap. 6)

Methods for Preventing Voltage Collapse

Enhanced Artificial Neural Networks Using Complex Numbers

Radial Distribution System Reconfiguration in the Presence of Distributed Generators

Comparison of Reference Compensating Current Estimation Techniques for Shunt Active Filter

ECE 2133 Electronic Circuits. Dept. of Electrical and Computer Engineering International Islamic University Malaysia

RBF NN Based Marine Diesel Engine Generator Modeling

Accelerating-Power Based Power System Stabilizers

Determination of Inrush Current to High Power Transformers using the LabVIEW Environment

Transcription:

Internatonal Conference on Advanced Electronc Scence and Technology (AEST 2016) Transformer wndng modal parameter dentfcaton based on poly-reference least-square complex frequency doman method Yanng L 1, a, Hong Yu2 and Xaoyan Zhu1 1 North Chna Electrc Power Unversty, Baodng, Chna Yunnan Power Grd Electrc Power Research Insttute Co., Ltd.,kunmng, Chna 2 Abstract. The modal parameter of transformer wndng, as drect reflectons of mechancal performance, s an mportant theoretcal foundaton n the feld of transformer manufacturng and detecton of wndng condton based on vbraton. To dentfy the modal parameters accurately, a modal experment on a 10 kv transformer wndng was conducted n ths paper. A PolyMAX method to dentfy wndng natural frequency and dampng rato was proposed usng the experment to dentfy transformer wndng modal parameters. And t uses modal confdence crteron to verfy. The frst four order modal parameters, ncludng natural frequency, dampng rato were extracted from the vbraton sgnal. The results demonstrate that all the natural frequences are far from two tmes of exctng frequency, whch ndcates that the desgn s approprate. The parameters verfes by MAC, whch verfes the effectveness of the proposed method n dentfyng modal parameters of transformer wndngs. Keywords: transformer; wndng; modal parameters; PolyMAX. 1 Introducton Transformer s one of the most mportant equpment n power system, and the stablty of power system s very mportant to the safety of power system. In normal operaton of the transformer, the wndngs are excted by the electrc power to generate vbraton. If the natural frequency of the transformer wndng s close to the exctaton frequency of the axal electrc power, the wndng wll produce a resonance phenomenon. Therefore, t s mportant to dentfy the natural frequency of the wndng, whch s of great sgnfcance to the desgn and manufacture of the transformer wth the natural frequency of the wndng. In addton, n the long run or short crcut mpact, the wndng may appear loose or deformaton, and the wndng mode parameters can be accurately dentfed and explored. At present, the analyss of the nherent characterstcs of the transformer wndngs s lmted to a smple model of the wndng. The model of sprng - pad s used to calculate the [1-2], but the model parameters are dffcult to be determned, and the smplfed model and the actual transformer wndng structure are qute dfferent. The fnte element modellng and smulaton analyss of the transformer wndng and the body mode of the transformer wndng and the body of L Yan, such as, has not been verfed. Wang Hongfang et al. Study the effect of [4] on the nonlnear vbraton characterstcs of a Correspondng author : hdlyanng@163.com 2016. The authors - Publshed by Atlants Press 796

wndngs, but the nherent characterstcs of the structure are dscussed. L Hongku et al. The exctaton vbraton of the transformer wndngs s carred out by [5]. But only the frst order natural frequency of the wndng s extracted, and the wndng dampng rato and vbraton mode are not gven. Because of the complex structure of transformer wndng, nonlnear strong, so the vbraton sgnal contans nose nterference. In the tme doman or frequency doman, the dentfcaton method of modal parameter dentfcaton method s lmted to the vbraton sgnal alone, and the dentfcaton accuracy s lmted, and s senstve to nose. In ths paper, the exctaton vbraton of a 10KV transformer wndng s excted, the PolyMAX method s used to dentfy the modal parameters of the transformer wndngs and the 4 order natural frequency and dampng rato are obtaned. Fnally, the MAC (modal confdence crteron) s used to verfy the dentfcaton results. 2 Basc prncple 2.1 PolyMAX modal dentfcaton method The PolyMax modal dentfcaton method s used n ths paper, also known as mult reference least square complex frequency doman method (Polyreference for complex frequency doman method), proposed by the Belgan B. Peeters and H. van der Auweraer professor n 2004, and s based on the least square estmaton theory and the mpulse response functon and the poles and resdues between the complex exponental relatonshp, and use the matrx of frequency response functon as a fttng functon n the frequency doman analyss method. The method s frst by sngle pont exctaton mult-pont vbraton pckup system response measurement ponts and between the exctatons s obtaned the frequency response functon, by the frequency response functon and through the standard algorthm. Fnally, we get requred for the model parameters. At the same tme, the method s not lmted by the sze of the dampng, the ntensty of the mode and the nterference of the nose. Because the pulse functon contans all the nformaton of the mode, t can be used to dentfy all the modal parameters of the structure system. Basc calculaton procedures are as follows: H ( ω) computng test system transfer functon H ( ω) obtaned based on the measured nput sgnals xt () and yt (),as follows: [ H ( ω)] l m [ B( ω)] = [ A( ω)] l m m m (1) Wheren, B( ω) s a frequency doman representaton of the output sgnal yt (); A( ω) to form a frequency doman output sgnal xt (); m s the number of the reference channel nput stmulus; l s the total number of channels n response to an output. If B( ω) and A( ω) can be expressed as follows: r [ B( ω)] = Z [ β ] p r= 0 p r= 0 r l m r [ A( ω)] = Z [ α r] m m (2) (3) For any frequency ω, lsts of equatons as represented by (1) can be measured accordng k. If we take a dfferent frequency, a suffcent number of lsts (over determned) equaton, t to [ H ( ω )] 797

can polynomal coeffcents [ β ] and[ α ]( r = 0,1,, p) obtaned by the least squares estmaton of r r the numerator and denomnator of the matrx to be determned. [ β ], [ α ] are real-valued coeffcents of each element, they are set to the complex s possble. r r [ α ] appears on the denomnator of the equaton s the nonlnear parameter dentfcaton problem. r However, a certan degree of lnearzaton, the varable of the lnear least-squares estmaton problem. α ( r = 0,1,, N 1) n the denomnator polynomal coeffcents based on known and gven the r α = I, and construct ts adjont matrx for egenvalue decomposton, we can get the system poles N and modal partcpaton factors, as follows: 0 I 0 0 0 0 0 0 V = VΛ 0 0 0 I T T T T α0 α1 αn 2 αn 1 (4) Λ characterstc value for the matrx (dagonal matrx), dagonal elements λ ( = 1, 2,, pn) as follow: λ = e p t or * p t e (5) * p, p = σ ± jω d (6) Modal dampng rato as follows: ξ σ σ = = ω 2 2 σ + ωd (7) In all matrces obtaned polynomal coeffcents A and B, the frequency doman usng the least squares method the modal shape, the fttng functon as follows: ψ l ψ l L H( ω) = ( + ) + U N T * H r r r r * 2 r= 1 jω pr jω pr ω (8) 2.2 Verfcaton based on MAC After the test analyss s completed n order to prevent the ntroducton of false mode or the loss of real mode, t s necessary to verfy the modal parameters n order to use MAC to verfy the applcaton. MAC s the correlaton coeffcent of two modal vectors ψ1 and ψ 2, as follows: ψψ 1 2 = 1 MAC = L L ( ψψ )( ψψ ) L 1 1 2 2 = 1 = 1 (9) The modal vector s ndependent of each other, and the MAC value between the two real modes should be close to 0. 798

3 Experment and smulaton 3.1 Expermental descrpton In ths paper, the hgh voltage wndng of a power transformer wth S9-630/10, the rated voltage of 10kV, and the connectng group are Y0. In order to be able to smulate the mechancal characterstcs of hgh voltage power transformer wndng, the wndng s the same as that of the power transformer. The transformer s carred out wth a KISTLER8763 acceleraton sensor, and the senstvty s 2mV/g, whch s arranged on the wndng surface by the metal clamp. At the top of the wndng, the B&K4808 electromagnetc excter s used to stmulate the wndng. The exctaton sgnal s whte nose sgnal, and the bandwdth s 20 khz, whch s provded by the dgtal sgnal generator and connected to the electromagnetc excter by the power amplfer. The electromagnetc excter s used to measure the exctaton sgnal and the expermental scene s shown n Fg. 1. Fgure 1. Overall vew of transformer wndng modal experment The expermental apparatus ncludes sgnal acquston and analyss system, charge amplfer, vbraton acceleraton sensor, vbraton sensor, power amplfer and sgnal generator. Fgure 2. Sgnal generator, power amplfer and sgnal acquston system 3.2 Modal test of transformer wndng In the nteracton of the leakage magnetc feld and current of the transformer wndngs, the electrc power wll be changed wth the pulse of current and the magnetc feld strength, whch causes the wndng vbraton. When the sudden short crcut, the transformer wndng s easy to produce the loose and the deformaton under the huge short-crcut current. After the wndng loose and the deformaton wll cause a greater mbalance electrc exctaton force, and form a vcous crcle, the wndng s damaged, even lead to serous accdent of the transformer. Therefore, the capacty of the transformer s the key to the desgn of large capacty transformer. Short crcut force can be classfed as the axal force, radal to force and crcumferntal force (of spral wndng), the role of axal force makes the wndng n the axal drecton s compressed, resultng n axal dsplacement and axal mechancal vbraton; radal force makes the wndng s compressed, wndng stretch; crcumferntal force the wndng generates torson effect. In the course 799

of the short crcut, the transformer wndngs are generated under the acton of short crcut force. The dynamc force n the wndng s not only short of short crcut electromagnetc force, but also the frcton force of nerta force, elastc force and dsplacement of the wndng. The dynamc process s very complcated, whch s nfluenced by the electromagnetc, mechancal, materals and other factors. In general, the calculaton of the dynamc process of the transformer wndng, the axal movement of the transformer wndng under the axal force and the radal movement of the transformer under the acton of the radal drecton. Therefore, ths study tests the axal mode and the radal drecton of the transformer wndng. 3.2.1 Transformer wndng axal modal test The purpose of the test s to measure the axal vbraton of the wndng, and the natural frequency and the correspondng vbraton mode of the wndng of the transformer are obtaned. In the experment, the whte nose s used to motvate, and there are 20 vbraton acceleraton sensor measurng ponts. Fg. 3 s the locaton of the vbraton sensor. Fgure 3. Vbraton sensor locaton dagram The exctng pont s located n the promnent part of the concave steel. The excter adopts elastc suspenson, vertcal exctaton, exctaton rod ends a force sensor. Fg. 4 for the excter placed n knd. Fgure 4. The vbrator s placed n the physcal dagram Fgure 5. The A phase wndng frequency response functon 800

Fg. 5 s a phase wndng axal expermental modal testng the vbraton sgnal frequency response functon s the result of the superposton, usng a sem logarthmc coordnate representaton. You can see from the fgure, the frequency response functon have obvous peak superposton results n 229Hz, 322Hz, 501Hz. Usng the PolyMAX method to dentfy the natural frequency of the A phase wndng, the frst 4 order natural frequency of the wndng s obtaned as shown n Table 1. Tab. 2 and Fg. 6 gves the MAC values of the natural frequences of the dfferent frequences and the MAC dagrams. By chart shows, the A phase wndng under normal operatng condtons, the frst 4 order of the axal vbraton natural frequency dfference s larger, and are far away from the exctaton frequency of the wndng s 2 tmes that of 100Hz, the structure s more reasonable. At the same tme, the MAC value of the 2 order modes can be consdered to be effectve n extractng the natural frequency of the 4 order vbraton. Table 1. Four order natural frequency of wndng Order 1 2 3 4 Frequency(Hz) 230 322 510 724 Dampng rato(%) 1.71 0.62 1.70 1.54 Table 2. MAC values of the frst four order modes of the A phase wndng Order 1 2 3 4 1 1 0.06 0.04 0.06 2 0.06 1 0.02 0.02 3 0.04 0.02 1 0.01 4 0.06 0.02 0.01 1 Fgure 6. MAC dagram of the frst four order modes of A phase wndng 3.2.2 Transformer wndng ampltude mode test The purpose of the test s to measure the frequency and vbraton of the transformer wndngs and the natural frequency and the correspondng vbraton mode of the transformer wndng. In the experment, the sne sweep exctaton s used to measure the 20 vbraton acceleraton sensor. 801

Fgure 7. Overall frequency response dagram of wndng The superposton of Fg. 7 for the A phase wndng radal modal experment test the vbraton sgnal obtaned by frequency response functon. You can see from the fgure, the frequency response functon have obvous peak superposton results n 6.75Hz and 20Hz. Usng the PolyMAX method to dentfy the natural frequency of the A phase wndng, the frst 3 order natural frequency of the wndng s obtaned as shown n Table 3. Table 4 and fgure 8 gves the MAC values of the natural frequences of the dfferent frequences and the MAC dagrams. By chart shows, A phase wndng under normal operatng condtons, the frst 2 order of the natural frequency of the frequency dfference s larger, and are far away from the exctaton frequency of the wndng s 2 tmes that of 100Hz, the structure s more reasonable. At the same tme, the MAC value of the 2 order modes can be consdered to be effectve n extractng the natural frequency of the frst 2 order vbraton. Table 3. Two order natural frequency of wndng Order 1 2 Natural frequency(hz) 6.75 20 Dampng rato(%) 5.22 4.79 Table 4. MAC values of the two order mode of the wndng Order 1 2 1 1 0.32 2 0.32 1 Fgure 8. MAC dagram of the frst two order mode of wndng 4 Concluson In ths paper, the vbraton test of the transformer wndngs s carred out. The PolyMAX method s used to dentfy the modal parameters of the transformer wndngs. The 4 order natural frequency and 802

dampng rato of the transformer wndngs are extracted. The MAC value and the MAC chart are ntroduced. Expermental results show that the modal parameter dentfcaton based on PolyMAX method has good nose resstance, and ts accuracy s hgher than that of the tradtonal frequency doman dentfcaton method. At the same tme, because of the expermental transformer wndng s a pe col structure, and large power transformer wndng structure s same, so ths paper s also used n 220kV and 110kV and other hgh voltage grade power transformer wndng modal parameters dentfcaton, dentfcaton results can provde bass for transformer desgn, manufacture and vbraton based wndng state detecton. References 1. SHAO Yuyng, XU jan, RAO Zhush, et al. Applcaton reseach of detecton transformer wndngs deformaton based on vbraton frequency response analyss[j]. Journal of Shang ha Jaotong Unversty, 44 (9):1223-1228, (2010). 2. GUO jan, LIN Heyun, XU Zhong, et al. Analyss of axal stablty of power transformaton wndngs usng fnte element [J]. Hgh Voltage Engneerng, 33 (11):209-212, (2007). 3. LI Yan, ZHOU We, JING Yongteng, et al. Axal vbraton analyss of transformer actve part under short crcuts [J]. Advanced technology of Electrcal Engneerng and Energy, 31(3):49-53, (2012). 4. WANG Hongfang, WANG Naqng, LI Tongsheng. Axal vbraton equvalent one degree analyss of power transformer wndng under short crcut [J]. Transformer of Chna Electrotechncal Socety, 15 (5):39-41, (2000). 5. LI Hongku, LI Yan. Axal vbraton modal analyss of transformer wndngs under dfferent level of precompresson[j] Electrc Machnes And Control, 14(8): 98-101,(2010). 6. Brat Peeters, Heram Van der Auweraer, Patrck Gullaume, et al. The PolyMAX frequency doman method: A new standard for modal parameter estmaton [J]. Shock and Vbraton, 11:95-409, (2004). 7. Carmona R, Hwang W L, Torresan B. Mult-rdge detecton and tme frequency reconstructon[j]. IEEE Transacton on Sgnal Processng, 1999(47):480-492. 8. Peeters B, Gullaume P. Automotve and aerospace applcatons of the LMS PolyMAX modal parameter estmaton method[c]//proceedngs of the 22th Internatonal Modal Analyss Conference,Dearborn,USA,January,(2004). 9. Gullaume P, Verboven P, Vanlandut S, et al. A poly-reference mplementaton of the leastsquares complex frequency doman estmator[c]//proceedngs of the 21th Internatonal Modal Analyss Conference. Kssmmee, USA, February, (2003). 10. [10] GENG Chao, WANG Fenghua, HUANG Hua, et al. Transformer Wndng Modal Parameter Identfcaton Based on Complex Morlet Wavelet Transform [J]. Journal of Shangha Jaotong Unversty, 47 (12):1981-1986, (2013). 803