Hi there, I am implementing an inverter loss simulation. I want to scope the switch and diode power loss in PLECS. In Simulink, I used a PLECS Circuit block as shown in the picture. I am sending my PWM signals to the inverter inside the PLECS block and get the line to line output voltage.
The load is modeled as a lookup table in Simulink. Since the load is not physical, there’s no current going through the switches in PLECS, thus all the losses are 0. How can I fix this?
Thank you very much for your time!
I think the best alternative is to simply replicate the above Simulink functionality in PLECS. These are all basic PLECS components as well, including he 3D LUTs.
If you must use Simulink, have you tried using voltmeters and controlled current sources to interface between Simulink and PLECS? You also need a voltage source on the DC side in both cases.
Thank you Bryan for taking your time answer my question. The issue with the LUT is that it causes no current flow in the inverters since it’s not a physical model.
Please take a look and see how rotor reference frame machine models are implemented. Here controlled voltage sources and controlled current sources are used to interface the machine equations with the circuit. Although it’s not a physical model, current flows because you are injecting it into the circuit. I modified the default PMSM model in PLECS so that you can see everything at one level. Does this make sense? This is a common modeling technique. Small note, but you might have to add three phase-to-phase resistances at the stator output for numerical reasons. If the resistors are reasonably large the power loss will be small.
Also, have you taken a look at the “Non-Excited Synchronous Machine” model? It also implements a LUT based approach similar to what you are doing here. However, instead of using the above interface approach it implements a Voltage Behind Reactance approach, where the interface to the inverter is electrical (hence it’s called the Voltage Behind Reactance).
