Performance of Inverter System
Forecast the Performance of Inverter Systems with multivariate inputs
ACS; RWTH
License
Mozilla Public License 2.0 (MPL-2.0)
Main Characteristic
A machine learning model for forecasting weekly inverter performance based on prior multivariate bi-weekly readings.
The model is developed with Tensorflow.
This solution is part of the I-NERGY project.
Research areas
Physical AI
Technical Categories
AI services
Business Categories
Energy
Last updated
21.09.2022 - 20:36
Detailed Description
This solution leverages ConvLSTM to encode latent information from the irradiance, ambient temperarure, nominal energy rating , and the internal inverter temperatures to forecast weekly performance values. The solution was entirely developed in python and Tensorflow 2.10.