能讲讲emtp中各种传输线模型的区别吗

monkeydl

能讲讲吗？详细介绍一下Lcc，或者给个相关的资料

dongchch

EMTP Theory第3，4，5章，尤其是第4章！

yizhangsjtu

I remembered I explained this question long time ago on PSCAD board.  For the convenience for those who are interested in EMTP theory or application, I can talk about it again.

  l, h' ^0 v+ B6 s5 _9 q9 A& i7 QThe transmission line (TLINE) model can be PI section, Bergeron, frequency dependent modal domain (J Marti) line  and Cable (L Marti), and phase domain frequency dependent model (which is not covered in EMTP theory book, but belong to TLINE model).
1 b2 s: g6 }: ]+ B4 K
PI is good for short line less than 15 KM.
Bergeron is a traveling wave model but not considering the frequency dependent effects.
J Marti model is a good approximation for over head line. It fix the transformation matrix at a frequency but consider the frequency effects on RLC.
( w5 X6 Q+ K0 l; Y  D" _1 C: @) }L Marti model improve the J Marti model by varying the transformation matrix. However it only works good for cable. There is numerical trouble with the over head line.
; a6 ~* |1 G1 a/ m) dPhase domain model is better by avoiding the mode transformation. The remaining problem is no passivity cause by the rational fitting. This is a still open area welcome further research and investigation.

dongchch

关于电缆的模型可以参考我和ywhywh的讨论
https://tech.cepsc.com/thread-38403-1-2.html

yizhangsjtu

I would like to comment on your discussion on  https://tech.cepsc.com/thread-38403-1-2.html, which is attached below.
$ b9 \5 b: x8 H8 F8 @9 o
2 O( b9 ]0 R# M7 @: ELuiz Marti is younger brother of Jose Marti, they are both students of Prof. Dommel in UBC. ATP has a problem that is all the development is volunteer based. There is no sufficient fund to support the work of ATP. The work of ATP was support by the US government, that is tax payer of US, so it is free to everyone. DCG is a supported by EPRI so it still goes forward. DCG was lead by Ontario Hydro, where L Marti is a leading engineer in the EMTP group. In fact there is another EMTP version called UBC version which is lead by J Marti.

% h- e; S7 d/ Q* c7 o2 k1 D: SI did not read the paper of Noda, since I did not used ATP for a long while. I think Noda's method must be a type of phase domain model which avoid the frequency dependent transformation matrix.  Gustavson's work is more popular in this area. In fact he is still work on this thing now. There is two difficulties now on the phase domain model, one is the accurate representation of DC resistor, the other is the non passivity of the fitting. I met Gastvson earlier this year and had a short discussion on this issue. He himself is trying to develop better fitting algorithms, which naturally can reduce the probability of non passivity. Current passivity enforcement is used once the problem appears, however, it bring tolerance. Chance is there in future a good fitting approach will be found and it reduce the chance of non passivity to a very low.

There are some tip to use the fitting, such as you can vary the weighting factor on the frequency. For DC, it is good to set the working frequency to 0.5. Don't use 0, it will bring numerical problems. I expect those phase domain line model will give use the option to choose the weight factor on the fitting. You may play several time and get a good number.

关于电缆的模型可以参考我和ywhywh的讨论
! K/ F' e+ ^0 lhttps://tech.cepsc.com/thread-38403-1-2.html
dongchch 发表于 2009-6-29 04:49

                               
登录/注册后可看大图

+ C3 x0 t  v) u# |  n其实，我感觉ATP在发展电缆模型上有所失策。L.Marti（不知道和JMarti是什么关系）在88年的时候其实已经提出了适合电缆模拟的模型，不过这个模型被EMTP的另一个版本DCG/EPRI所采用，ATP里并没有采用，一直到了96年，Noda （Ametani的学生，Ametani是cable constants/parameter 的作者，EMTP的牛人之一）提出了直接在时域计算的Noda模型，不知道究竟是因为这个模型很先进，还是因为Ametani教授的权威地位，反正ATP 决定引进这个模型了，没过3年，Gustava又提出了计算更稳定的模型，这个模型马上就被pscad所引入。ATP却一直就只保留着Noda模型了，操作比较复杂，据说计算有时也不稳定，没用过，但是感觉影响力不如那两种模型

dongchch

I would like to comment on your discussion on  https://tech.cepsc.com/thread-38403-1-2.html, which is attached below.
" q+ _; Z8 h) t/ s: T' Q. A
; R& Q! P  V. V0 T' ^& CLuiz Marti is younger brother of Jose Marti, they are both students of Prof. Domm ...
yizhangsjtu 发表于 2009-6-29 09:31

                               
登录/注册后可看大图

) \' X! R0 N; J/ h/ J2 G6 F0 k/ a非常感谢张老师的详细的讲解！关于电缆模型我还有个疑问，如您所提到的，fitting的算法似乎一直都是模型改进的热点，可是有没有把电缆半导体屏蔽层(semiconductor)加入到模型里的研究呢，半导体屏蔽层对波的传输影响很大，这一点已达成共识，为什么却没有一个程序考虑加入这个模型呢？我看到的只有Gustava提出的间接考虑semiconductor的办法，但好像有人实验验证过(很久以前读到的，忘了出处了)，这种处理果并不理想，您有何高见? 欢迎讨论！

yizhangsjtu

The current approach to model the insulator is by based on the resistivity  and relative permeability of the material. People may think to represent the semi-conductor material as something in between the conductor and insulator. As far as I know, this method is not available yet. Chance may be there to take some new progress in the material science or electronics science area, then improve the current modelling of cable.
; V1 X( C8 Y3 X1 M
( J; {0 Q1 d: [! ~7 eI would talk to Bejorn Gustavson about this issue if I see him on the PES General Meeting next month. I will keep you updated if anything news turns out.

dongchch

7# yizhangsjtu
受益不少，非常感谢您的回答！

ywhywh

non passivity of the fitting是什么意思呢？
) T" M. U' C( G) x7 U- m1 m0 _我对fitting的理解就是用几阶的方程式近似替代模型进行计算，就像可以用一个无穷级数代替指数函数一样。

liuyiqing

路过，学习一下！