Outlook

The question of how the solar module optimizer behaves with a real solar cell under partial shading could not be investigated in this work. However, the PV emulator shows very similar behavior compared to a real solar cell. Here, the experimental trial with two same solar cells and the module optimizers but different algorithms under environmental conditions could be another research project. In this case, data should be recorded on the SD card over 24 hours for later comparison.

The programmable solar module optimizer can be used as a platform for the development of more advanced algorithms. Here, additional sensors such as the temperature of the solar cell or weather data can be obtained via a WIFI-connection. Solar cells with a different technology such as thin-film or perovskite, which have a more complex characteristic curve to collect, are of interest. In addition, a connection to a AI could be established through the serial interface or WIFI, this would allow the analysis of much more complicated characteristics.

In addition, the PI controller can be adapted to the extent that the output voltage is no longer observed, but the output current, this would probably make the control loop usable again to thus realize a functioning charge controller, this could then charge a lead battery, for example. Here the output voltage would always be pulled to the $12V$ of the lead acid battery and the additional energy of the coil, results not in a higher output voltage, but in a larger charge current. Different MPP charge controller algorithms could be investigated.

Furthermore, the Mircopython approach allows the platform to be used for teaching, because it greatly facilitates programming, and through this, a better understanding of VI-characteristics can be given to MPP and PV.