AI-based Online Warning and Lifetime Prediction System for Transformers, Motors, and Generators

Purpose and goal: ** Denna text är maskinöversatt **

In the work, we have successfully produced a PoC that can predict PD and thus future health for electrical systems and components. In the work, we successfully developed the associated hardware component, software and investigated the best application with the associated customer benefit. We consider ourselves to have received encouraging results that seem promising for continued development of the innovation. With the project completed, we are now ready for further development and testing on a larger scale.

Expected results and effects: ** Denna text är maskinöversatt **

After completing project we have now created greater awareness of our product and its potential. In our work, we have demonstrated and highlighted the value our technology can create for individual actors but also for society at large. Approach and implementation: ** Denna text är maskinöversatt **

The work has been led by WindingsAI where we have carried out the following activities Data collection and hardware assembly, ML-based PD signal filtering, ML-based feature extraction from PD, ML-based system for early warning and prediction of equipment life and efficiency, Testing and improvement and early commercialization and business development.

We realized in our work how important it is to carefully carry out the fundamental steps of measurement, filtering and feature extraction in order to achieve good results for the prediction.