Iterpolation shows wrong results

Hello

I have a major problem with my results that I can't seem to find its reason

My model beginns with an electrical loads which is supplied by a synchronous generator. the generator gets its power via an asynchronous motor (galvanic isolation of an island grid)

and this is the electrical load, connected to the generator, which was interpolated using the interp1 spline method and goes to a 1D lookup table

the values of the 1D table are measurment points each 30s and where interpolated to points each 10s (1 sample point in matlab=10s --> total of 90 sample point) and working with a finer interpolation would last so long

ah shown the plot looks smooth and the torque od the induction motor looks fine so far

the problem is the speed which has these semi-intervall between each two points (like between 18 and 19)

what I would like to achieve is what I've drawn with a blue line

I guess it's a problem related to the interpolation right? I don't know why I have these semi-Intervall, and the moving average block didn't help

the lookup table has the following setting

I have the following questions:

• is it a solver problem? I used fixed step discrete size 1e-5
• is it an interpolation problem?
• if fixing the model isn't possible, is there any other war?

I'd be gratfull for any suggestions or ideas

regards,

[vidyesh]

Can you attach the simulink model?

Answer 1

[K]

thank you

the problem was clearly a step size mistake. I implemented each measurment point to a natural number (1,2,3...) which made the plot have these semi interval. when I implemented each measurment point in step size of 0.1 (like 1.1;1.2;1.3...) I got better plots, which takes me to my second question (that I posted after):

how do I know the step size of a discrete model? I used all the same parameters of the Electric engine dyno model (https://de.mathworks.com/help/sps/ug/electric-engine-dyno.html?searchHighlight=dyno&s_tid=srchtitle_support_results_1_dyno)

I think it has something to do with the "Control sample time " and "Fundamental sample time " for FOC, to which I didn't find answers in documentation

new step size with which I implemented the measurments value is 0,1 which happens to be the ratio (Fundamental sample time / Control sample time).

it also worked with step size 0,2 which is (Fundamental sample time*10*switch frequency) so I guess these are the parameters that defines the step size.

can you help me understand the step size of the engine dyno model?

regards

Answer 2

[Sabin]

Fundamental sample time defines the sample rate at which the system is executed. The control sample time is the rate at which the control algorithm is executed. When we discretize a system, we should ensure we have a small enough sample time to capture the system dynamics. When we have PWM signals we also need to oversample the system to ensure we capture all the edges with relatively good accuracy.