thank you, The conclusion of experimental results that are close to the results of the practical aspect and is done in more than one way, such as MATLAB, which is the most famous in drawing conclusions

To control, monitor, and supervise the signals that come out from each device applied in the HIL.

The HIL setups allow us to test a real piece of hardware (hence the name - hardware-in-the-loop) with its pertaining features and operational modes.

Specifically, in a controller-hardware-in-the-loop (C-HIL, just one type of HIL setup) setup we can test how the controller with an embedded code is going to behave once deployed in an actual device. What is hard to model or do in e.g. MATLAB in this regard is the following:

- Controller capacities utilization - in various forms! (memory overrun, interrupts execution, etc.),

- Coordination and execution of different control layers (especially if they are executed on different controllers/control platforms),

- Communication devices and systems, overlaying the power electronics system (how are you going to model dynamics associated with the CAN communication protocol, for example, used to exchange the references between the controllers?),

- Integration with other real-time executed tools (for example, flight simulator for testing more electric airplanes),

- etc.

So to abstract things - Doing tests on a certain piece of hardware in real-time allows us to connect these devices to other devices and processes that are also executed in real-time (and sometimes can not be easily executed in non-real-time context) and to examine many features that are simply hard to model, or sometimes meaningless, in an offline simulation.

On the other hand, and more broadly speaking, unlike MATLAB, the HIL setups allow us to model and test:

- Medium/large complexity and size systems (for example, microgrids are extremely hard to model in detail since the model runtime tends to increase exponentially),

- High-power/high-voltage devices (in a power-hardware-in-the-loop, P-HIL, setup we can test the behavior of the real protection device in various network situations, such as short-circuits, loss of phase, etc. - this is next to impossible to do in offline simulation),

- etc.

This all being said, the offline simulation tools (constituting model-in-the-loop, MIL, setup) are here to stay. They are still great for initial tests, to put it in simple terms, and simulation tool providers are constantly adding new useful features, but the HIL setups have closed the serious gap that existed between the offline simulations and the tests on actual devices (prototypes or commercially available devices). It is safer than the real hardware, but faster and 'more precise' (again serious simplification) than the models executed in offline simulations.

You can check the following two documents for further details if you wish.

Chapter Advances and Prospects in Distributed Generation Sources Dig...

https://www.mdpi.com/2673-4931/11/1/2/htm

Article Advances in Distributed Generation Source Digital Twins Design

There are several reasons to do HIL tests instead of offline simulations only.

1) strictly speaking any simulation model for the device being tested is a simplification of the actual device to certain extent.

2) the implementation of the design of the device may not be perfect and you need to test the actual device to make sure that the manufactured device is true to your design, which can only be done in a HIL test setup.

3) the real time simulator runs much much faster than the offline simulation in most cases which make testing process more efficient.