外文翻譯--基于δ調(diào)制和pwm控制異步電機(jī)的時(shí)間-頻率分析（英文）

1、Serbia data compression; image coding and analysis; communications; speech andacoustic signal processing, and medical signal processing. The time-frequency plane is a rich feature space for analyzing the signal's at

2、tributes. A TFR of a multi-Daniel N. Trip is with the Department of Electronics, University of Oradea, University Str. 1, 410087, Oradea, Romania (phone/fax: +40- 259-408191; e-mail: dtrip@uoradea.ro). Cornelia E. Gordan

3、 is with the Department of Electronics, University of Oradea, Romania (phone/fax: +40-259-408191; e-mail: cgordan(uoradea.ro). Mircea I. Gordan is with the Department of Electrical Engineering, University of Oradea, Roma

4、nia (+40259-408191; e-mail: mgordanguoradea.ro). Adrian Schiop is with the Department of Electronics, University of Oradea, Romania (e-mail: aschiop@uoradea.ro). Romulus Reiz is with the Department of Electronics, Univer

5、sity of Oradea, Romania (e-mail: rreiz@uoradea.ro)component non-stationary signal consists of sets of ridges,the orientations and widths of which characterize thesignal. For example, once computed, time-frequency images

6、can be processed using edge detection and other image processing algorithms to automatically determinethe ridge parameters. The STFT and continuous wavelet transform have been suggested for the first, image generation st

7、ep of the feature extraction procedure. In thispaper, we will explore the advantages of using TFRs for analyzing the signals that appear in a system used for Delta Modulation and PWM Control of the Asynchronous Motors.Bi

8、linear time-frequency distributions, offer a wide range of methods designed for the analysis of nonstationary signals. Nevertheless, a critical point of these methods is their readability, which means both a good concent

9、ration of the signal components and no misleadinginterference terms. Some efforts have been made recentlyin that direction, and in particular a general methodologyreferred to as reassignment. TFRs are two dimensionalfunc

10、tions of time t and frequency f that indicate how the frequency content of a signal x changes over time. Thesimplest TFR is the spectrogram, the squared magnitude of the STFT:(1)The classical time-frequency resolution tr

11、adeoff of thespectrogram, which is controlled by the analysis window w, has prompted the development of more advancedbilinear TFRs, including the Wigner distribution.w (t,f) J (t+ TJ *jt T) ;ji2`ftdT (2)This TFR can be i

12、nterpreted as a short-time Fourier transform with the window matched to the signal. Whilethe Wigner distribution is highly concentrated, due to itsnonlinearity it generates cross-components and is verysensitive to noise.

13、 The spectrogram and Wignerdistribution both belong to Cohen's class of TFRs. The Wigner distribution can be interpreted as the central, generating member of this class, with each Cohen's class TFR C obtained via

14、 the two-dimensional correlation,C x (tj f) = | Wx (c, v)+(c - t, v - f)d-cdv (3)1-4244-0049-X/05/$20.00 (C2005 IEEEEUROCON 2005i S,(t,f)= x(-c)o)*(-c-t)e-““'d-c1586programming environment. One can see that the contr

15、olvoltage u, look like a pulse width modulated signal.IlllllIu t) [V]0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 t[S]200 0.002 0.004 0 006 0 008 0.01 0E012 0.014 0.016 0.018 0.02 t[S]u r (t) [V]0 0E002 0E

16、004 0E006 0 008 0E01 0E012 0.014 0.016 0.018 0.02 t[S]Fig.3. Waveforms for the delta modulatorIII. ELECTRIC DRIVE OF THE THREE-PHASE ASYNCHRONOUS MOTORThe block diagram of the proposed system is presentedin the Fig.4. On

17、e can see one 50 or 60Hz sinusoidal waveform signal to the input of each delta modulator.ThnWae phIas shif bewe thes signals is 1200 Due tothSprainpincpe of te1Delta modulaGtoBr,dg PWM __Sine ~ ~ ~ ~ ~ ~~~~~. ia,e bet o

18、oWrsae~~~~~~~~~~~~r.-three phase power inverter. TheDload.o power T e Workspa-e3Fig.4. Block diagram of the control systemThe phase shift between these signals is 120m. Due tothe operation principle of the delta modulato

19、r, PWM command signals will result. These signals are used to command the full bridge of athree phase power inverter. The load of the power inverteris in this case a three phase asynchronous motor. The motor speed can be

20、 modified if the parameters of the delta modulator blocks or the inverter's supply voltage are changed. In the Fig.5 are presented the simulation results of the control system during start transition time and at a lo

21、ad variation encountered after Is: ia - stator current, com- rotor speed, tem - electromagnetic torque and T - loadtorque.141llSlll j “.l I.1sl..I,, ., l, K.,I. 11 Ij ,, Iua... 11 ,L LILIs LJ.1A LJAhl jII L411Ld iLL2 LUA

22、 J 1 A LiJA111I 1s I,,l 11LL Ill~Parameters of the asynchronous motor are the same asin the model given in the Power Blockset.IV. TIME-FREQUENCY ANALYSISThe analysis of non-stationary signals is of interest in many field

23、s but also in power electronics. The study ofthese phenomena typically implies the use of a time- frequency (TF) analysis [10], [11], because it can give anoverall view of the behavior of non-stationary signals by means

24、of the so-called time-varying spectrum P(t, f ) (orP(t, u)). This spectrum is defined in the time-frequency plane (t-J), In the case of time-frequency analysis, the main problem is that of obtaining a good frequencyresol

25、ution and 'readability' of the spectrum, together withthe possibility of highlighting its evolution over time. The time-frequency representations of a signal contain very important information concerning the regi

26、ons fromthe time-frequency plane where the signal's energy is maximum. It was demonstrated that the ridges of the module of any time-frequency representation correspondto the maximum values of the signal's energy

27、, which formthe skeleton of the analysed transform. These maxims arelocalised around the instantaneous frequency (IF) of thesignal, which means that the detection of the ridges offersthe possibility to estimate the signa

28、l's signature. The importance of the instantaneous frequency concept stems from the fact that in many applications the signal analyst is confronted with the task of processing signals whosespectral characteristics (i

29、n particular the frequency of thespectral peaks) are varying with time. These signals areoften referred to as “nonstationary“. For these signals, theridge is an important characteristic, because it is a time- varying par

30、ameter that defines the location of the signal'sspectral peak as it varies with time. In our analysis we used the Gabor time-frequency representation, because it has the best behavior and it presents a reduced amount

31、 ofinterference terms. The analyzed signals were statorcurrent, rotor speed and electromagnetic torque. These are non-stationary signals because some of their spectral components are varying in time. The rotor speed sign

32、al presents a sudden amplitude modification that can be observed when a load torque isapplied. The time-frequency representation of the signal shows that the frequency content remains unchanged intime. So, there aren'

33、;t new spectral components thatappear. The only parameter that changes is the amplitude of the spectral components, as can be observed in Fig.6. -1000 E E]5 1 15 2 2.5 t[s] WM 2000 2W-200 l 0 0. 1 1.5 2 2.5 t[s] tem 200-