[D]Enhancing Weather Forecast Accuracy: Exploring Regression Models with Multi-source Data Integration

I am currently working as a data scientist at a new energy startup, mainly responsible for predicting photovoltaic power generation every 15 minutes for the next day. The key data relied upon are weather forecasts, especially the predicted solar irradiance values. Currently, we have data from five numerical weather forecasts, which include fields such as irradiance, temperature, and humidity. The accuracy of the forecasts varies among different data sources, and there are certain discrepancies with the actual weather. I am …

accuracy data data integration data scientist energy every five forecast integration key machinelearning next numerical power regression responsible solar source data startup the key values weather