Universität der Bundeswehr München - Institute of Flight Systems - AI-Lab
A workflow for the specific case of process monitoring data which contains cycles repeating over a long time is presented. The method relies on predefined features ideally used as health-indicators which are aggregated over the cycles. Anomaly detection, ...
In the context of AI4CZC project, this model forecasts the Montenegro net position (difference between generation and load) for the next 48h. The model was trained with data from 2019 to 2021 and tested on 2022 data. It performs with a MAE of 75.8 and a R...
A novel progressive training algorithm for learning Graph Neural Networks with Differential Privacy guarantees
A very short-term autoregressive model that predicts N step ahead
This model blends forecasts from two providers to improve accuracy
To provide accurate measures of Cross Zonal Capacity, a key parameter is Net Position (the difference between generation and load of a zone). Here we describe our supervised model to predict net position for the next 48h that achieve an MAE of 75.8 and a...