LS-PMSM stands for line-start permanent magnet synchronous machine. This is a type of electrical machine (i.e. generator or motor) that is a permanent-magnet synchronous type in normal operation, but which can be started by direct connection to the supply, utilising the rotor iron or some extra rotor conductors to make it come up to near synchronous speed as an induction (asynchronous) machine before lurching into synchronous operation. This gives the combined advantages of better efficiency than an induction machine (whose rotor carries currents and whose stator carries corresponding transformer currents) and greater simplicity and probably lower losses than an inverter-fed permanent magnet machine.
The subject of my MSc thesis (2001) was whether the replacement of common industrial induction motors with LS-PMSMs could be expected to change significantly the behaviour of the motors' interaction with the network.
The three documents that may be of interest are the thesis itself, the presentation that was made of it, and an article that was written then forgotten (largely because I still didn't believe at that time that conferences accepted such trivia -- now I know only too well). I apologise for the hideous typesetting and appearance. I was really new to computers at the time, having had my first one for less than a year (most previous work had been written up by hand, on the principle that computers waste a lot of time on diagrams etc.). I was not yet by that stage weaned off poor-quality bundled operating systems and word processors. Note in particular the almost touching fact that all figures from simulations have been carefully screen-grabbed and saved in some `paint' program; no ideas about eps or wmf export.
The simulation work was done in Simulink, and the
basic blocks can be found in the files in
the models directory.
I cannot give any assurance that they'd run in the current
version of Simulink -- remember they're from 2000/2001.
See the example system for how the blocks are used. Pure
simulink was used in preference to the, at that time at
least, appallingly arkward `power system blockset'.
Simulink had the appeal, over plain Matlab, of making the
computer-new user feel more lab-like in debugging problems --
sticking in a 'scope felt much better than poring over lines
of weird-looking computer program.
Incidentally: all files here are welcome to be used for anything
by anyone: public domain.
Page started: 2000-10-01
Last change: 2009-01-14