[Paper Review] Graph Based Power Flow Calculation for Energy Management System

Junjie Shi, Guangyi Liu+4 more|arXiv (Cornell University)|Oct 25, 2018

Microgrid Control and Optimization15 references4 citations

TL;DR

This paper proposes a graph-based power flow calculation method for Energy Management Systems (EMS) using graph databases and parallel computing to accelerate large-scale power system simulations. By leveraging nodal and hierarchical parallelism in a graph computing framework, the approach achieves significantly faster computation than traditional sequential methods, demonstrating strong potential for real-time EMS applications on practical systems.

ABSTRACT

Power flow calculation in EMS is required to accommodate a large and complex power system. To achieve a faster than real-time calculation, a graph based power flow calculation is proposed in this paper. Graph database and graph computing advantages in power system calculations are presented. A linear solver for power flow application is formulated and decomposed in nodal parallelism and hierarchical parallelism to fully utilize graph parallel computing capability. Comparison of the algorithm with traditional sequential programs shows significant benefits on computation efficiency. Case studies on practical large-scale systems provide supporting evidence that the new algorithm is promising for online computing for EMS.

Motivation & Objective

• To address the computational bottleneck in power flow calculations for large, complex power systems used in Energy Management Systems (EMS).
• To improve computation efficiency for real-time or faster-than-real-time EMS applications.
• To leverage graph database and graph computing paradigms to model and solve power flow problems more efficiently.
• To design a linear solver with nodal and hierarchical parallelism to fully exploit graph parallelism capabilities.

Proposed method

• The power system is modeled as a graph, with buses as nodes and branches as edges, enabling efficient representation and traversal.
• A linear power flow solver is formulated and decomposed into nodal and hierarchical parallelism to exploit graph computing architectures.
• Graph database technology is used to store and query power system network data, enhancing data access and scalability.
• The algorithm is implemented using a graph computing framework that supports distributed and parallel execution.
• The solver is designed to handle large-scale systems by distributing computation across multiple processing units based on system topology.
• The method integrates with existing EMS workflows by replacing traditional sequential solvers with a graph-parallel alternative.

Experimental results

Research questions

• RQ1 Can graph-based computation significantly improve the speed of power flow calculations in large-scale power systems?
• RQ2 How does nodal and hierarchical parallelism in a graph computing model affect solver performance compared to sequential methods?
• RQ3 To what extent can graph databases and graph computing frameworks enhance the scalability and efficiency of EMS power flow applications?
• RQ4 Does the proposed method maintain accuracy while achieving faster-than-real-time computation in practical EMS scenarios?

Key findings

• The graph-based power flow method achieves substantial speedup over traditional sequential solvers in large-scale system simulations.
• Nodal and hierarchical parallelism effectively exploit the computational power of graph processing frameworks, reducing solution time.
• Case studies on practical large-scale systems confirm the method's feasibility and efficiency for online EMS applications.
• The use of graph databases improves data management and query performance for complex power system networks.
• The algorithm maintains numerical accuracy while enabling faster-than-real-time computation, supporting real-time decision-making in EMS.
• The results demonstrate strong promise for deployment in real-world EMS environments requiring high-speed power flow solutions.