Day-Ahead-Prognose von Photovoltaik-Einspeisung mit SAP HANA Predictive Analysis Library

Abstract

The use of renewable energy is continuously increasing. As a result, accurate forecasting of photovoltaic power generation is becoming more important for energy suppliers and grid operators. Reliable day-ahead forecasts are essential for electricity trading and stable grid operation.

The aim of this bachelor’s thesis is to develop a prototype forecasting service for day-ahead photovoltaic power generation using the SAP HANA PAL. Historical photovoltaic generation data and weather data were collected, processed and integrated into an SAP HANA Cloud environment.

Different forecasting methods were reviewed, and the Hybrid Gradient Boosting Tree (HGBT) algorithm was selected for the implementation. Forecast accuracy was evaluated using the metrics MAE, RMSE, nMAE and nRMSE. In addition, a forecasting service was developed to automatically prepare the data, execute the model and provide the forecasting results within the SAP HANA platform.

The results indicate that the developed forecasting service provides a suitable basis for data-driven PV forecasting within an SAP-based system environment.

Volltext

Verfügbar auf Anfrage.