ESP32 Environmental Sensors and Solar-Voltage Forecasts

Overview

The source describes a low-cost ESP32 monitoring device with incremental solar-voltage forecasting. At each 15-minute interval it predicts the following 15-minute step. A daily curve comprises 96 sequential predictions; it is not a forecast issued 24 hours ahead.

Inspected version 3 files

The archive contains four CSVs and an Android APK. Ulm measurements have 481 rows and 6 columns, dated 17–22 May 2024. Zapopan training has 8,064 rows and 5 columns, covering 6 February–30 April 2025. May validation inputs have 2,976 rows and 5 columns; the corresponding measured/predicted voltage table has 2,976 rows and 3 columns. These validation tables describe the same timestamps and must not be counted as independent observations.

Limitations

The paper cautions that daytime heating of the enclosure biases its temperature sensor above ambient air temperature. The source also reports clock misalignment in the final three validation days. Firmware, a README and raw JSON are mentioned on the landing page but were not present in the downloaded archive. The APK was not executed.

Source and provenance

Dataset DOI: https://doi.org/10.17632/gvkc67gn5x.3. Related Electronics paper: https://doi.org/10.3390/electronics15173911.