Decentralized Control of DC–DC Converters for Redundant Onboard Power Supply Based on Artificial Swarm Intelligence Approaches

This paper applies basic swarm robotics concepts to the control of a redundant multi-source power supply system for an automotive onboard grid. The fundamental concepts and thoughts are explained to give a basic understanding of swarm robotics engineering, redundant vehicle power system structures, DC-DC converter modeling and control. A simplified example application is modeled and simulated as a first proof of concept. The results show the feasibility of the artificial swarm intelligence based concept. The numerical assessment shows significant improvements in dynamic control quality, bandwidth, stability, robustness and fault tolerance without the need for additional dedicated communication compared to a traditional single-source solution. In a second simulation, an adaptive and fault tolerant PWM carrier shift optimization is successfully realized. This demonstrates an even more advanced technical feature in addition to a pure high-stability voltage regulation and gives an outlook to the new possibilities opened up by applying swarm robotics concepts.