Power Converter Maintenance Optimization Using a Model-Based Digital Reliability Twin Paradigm

Abstract

Optimization of operations and maintenance activities in factories was estimated to have a global economic potential of 1.2 to 3.7 trillion USD by recent studies. Digital twins offer a framework to achieve such optimization by studying potential improvements in the virtual space before applying them to the real world. We studied the use of a digital twin based on a general model of system failure behaviour for maintenance optimization by combining existing methodologies into a general framework. Applying it to a real-world power converter use case, we identified either reactive or preventive maintenance to be more cost-effective depending on the operating conditions. This allowed to predict optimal maintenance for existing and future systems.