- 积分
- 9
- 回帖
- 0
- 西莫币
-
- 贡献
-
- 威望
-
- 存款
-
- 阅读权限
- 5
- 最后登录
- 1970-1-1
该用户从未签到
|
发表于 2013-9-20 11:45
|
显示全部楼层
来自: 马来西亚
There are actually two groups of parameter estimation method: one using dc signal, one using ac signal.
(refer Toliyat, H.A., Levi, E. & Raina, M., 2003. A review of RFO induction motor parameter estimation techniques. IEEE Trans. on Energy Conversion, 18(2), pp.271–283.)
For dc signal method, usually a step dc voltage is applied and the current response is measured. There parameters are identified using regression least square method.
This way, measurement of leg voltage using voltage sensor is easy.
For ac estimation method, the machine is supplied with pulse-width modulated (PWM) signal. As mentioned by angelguard and akljh himself/herself, voltage measurement is almost impossible due to the PWM nature of the signal. Hence, estimation from dc-link voltage and switching signals is more commonly used.
Beside the problem with dead-band, the voltage drop across the switch and diodes also have to be considered.
For dead time, I suggest reducing the switching frequency and/or reducing the dead time so that the effect is no longer obvious. Otherwise, try implementing dead time compensation method:
Jong-Woo Choi; Seung-Ki Sul, "Inverter output voltage synthesis using novel dead time compensation," Power Electronics, IEEE Transactions on , vol.11, no.2, pp.221,227, Mar 1996
doi: 10.1109/63.486169
|
|