# Guidelines and design considerations for designing a controller for a grid tie inverter?

I'm working on designing a controller for a grid tied inverter.

I'm not able to find a suitable reference or book that might help in this case.

What I have understood now is:

1. There must 2 loops (inner current and outer voltage loop) for better control
2. I have a basic understanding of modelling of inverter with LC filter using state space analysis.
3. Basic compensators except proportional resonant controller require DC for tracking. So direct-quadrature axes transformation is necessary. I understand the mathematics behind it.

I don't understand whether to use PI controller and tune it or manually design a controller using control theory.

I don't understand why there must be two loops and how to design the voltage compensator transfer function such that it gives current reference using voltage error.

Please help me with references or an explanation as I couldn't find any book that explains this in a mathematical manner.

This is similar to what Im trying to achieve. Where the ig and Vdc that is being given as a feedback to calculate error are direct-quadrature axis transformed.

• Please replace PR, DQ and TF with their English-language equivalents. You're OK with LC filter and PI controller, but it wouldn't hurt to clarify them. – TimWescott Jan 19 at 21:44
• forget about the voltage loop, a grid tie inverter will not be able to change the voltage on the grid, just control current. – Jasen Jan 19 at 21:46
• @Jasen Could you please elaborate. Cause I have read the same sentence in many literature. The purpose of the question was that I wasn't able to understand the mathematics behind it – Aravind Khumar Jan 19 at 21:49
• the grid is a voltage source, (very low impedance) to interface with that you don't want another voltage source, because a slight disagreement would require high currents to flow. If you build an AC current source and can keep it synchronised to the grid frequency then you can feed energy to the grid in a more predictable way. – Jasen Jan 19 at 21:54
• What did that sentence apply to? Were they talking about controlling the grid voltage and current, or some other voltage and current in the system? Were they talking about passive grid-tie inverters where you just dump as much power as you have onto the grid (which is what @Jasen is describing) or are you talking about active grid-tie inverters that participate in regulating the grid phase and voltage (which I can't point to in a reference, but which is becoming necessary as more and more power comes from distributed sources)? – TimWescott Jan 19 at 22:07