Distributed architecture for self-organising smart distribution systems

Abstract

Automation of emerging smart distribution grids is required to operate the grid efficiently and swiftly. This study draws a vision on grid automation with agent-based cyber-physical system integration to provide a truly distributed architecture. Furthermore, this study introduces a notion of self-organising smart distribution grid that promotes the grid capability to heal and organise itself in the best-suited topology without the intervention of a central operator. The proposed architecture comprises the system of bus agents (BAs) that emulate the given grid. This emulation is used by BAs to comprehend the grid conditions, switch location and compute their representative bus voltage and partial loss and to estimate the best organisation of the BAs as well as the grid. This study details the behavioural designing of BA that incorporates the functioning above. The proposed architecture also uses an event trigger approach to initiate grid organisation, which is showcased by case studies on IEEE 33 bus system. The results showcase the efficiency of the concept regarding solution accuracy with distributed computations; computational efficiency during contingencies; architecture performance under communication latency; and fault-tolerant characteristics of the proposed architecture.

Keywords

• distribution networks
• multi-agent systems
• smart power grids
• power grids
• distributed power generation
• self-organising smart distribution systems
• smart distribution grids
• grid automation
• agent-based cyber-physical system integration
• self-organising smart distribution grid
• IEEE 33 bus system
• distributed computations
• computational efficiency