The increasing complexity of modern electricity networks demands innovative computational approaches to power flow analysis, a critical task for ensuring grid stability and efficiency. Zeynab Kaseb, Matthias Moller, Peter Palensky, and Pedro P. Vergara, all from Delft University of Technology, present the first direct comparison of two promising quantum computing paradigms, gate-based computing and adiabatic computing, for solving these complex power flow equations. Their work adapts an existing adiabatic algorithm to a gate-based approach, allowing for a head-to-head assessment of performance using both a digital annealer and a gate-based quantum computer. The results demonstrate the potential of quantum algorithms to tackle the computational challenges facing modern power grids, offering valuable insights into the scalability and practical viability of each computing approach within the constraints of today’s Noisy Intermediate-Scale Quantum technology.
This research reformulates the power flow problem as a combinatorial optimization problem, specifically an Ising model, making it suitable for quantum solvers. The team compared the performance of the Quantum Approximate Optimization Algorithm, simulated on a classical computer, with that of actual quantum hardware from D-Wave and Fujitsu on a standard four-bus test system. Results demonstrate that quantum approaches can solve the power flow problem, although current hardware limitations prevent them from consistently outperforming classical methods.
This work highlights both the challenges and the potential of quantum computing for optimizing power systems. Traditional power flow analysis is essential for operating and planning power systems, but solving the associated equations can be computationally intensive, particularly for large systems. Researchers propose that quantum computing offers a potential solution by leveraging quantum algorithms to accelerate the optimization process. The core idea involves reformulating the power flow problem into an equivalent Ising model, a crucial step because quantum annealers and certain quantum algorithms are designed to solve such problems.
This transformation converts the power flow problem into a combinatorial optimization problem, more amenable to quantum solvers. The study investigates two main quantum computing approaches. The Quantum Approximate Optimization Algorithm is a gate-based algorithm, requiring a universal quantum computer, and was simulated on a classical computer in this work. Adiabatic quantum computing was explored using both the D-Wave system and Fujitsu’s Digital Annealer, both of which use quantum annealing to find the minimum energy state of an Ising problem. Experiments were conducted on a simplified four-bus power system, allowing for detailed analysis of solution accuracy and computational time.
The results show that both QAOA and adiabatic quantum computing can find feasible solutions to the power flow problem, with Fujitsu’s QIIO achieving the fastest iteration count. However, the authors acknowledge limitations in current quantum hardware, including qubit count, coherence time, and connectivity, which affect algorithm performance. The research team discusses the challenges of working with current quantum hardware and suggests that as quantum hardware improves, quantum computing could become a viable solution for solving large-scale power system optimization problems. Future research will focus on developing more efficient quantum algorithms, exploring different quantum hardware platforms, and investigating the use of quantum machine learning for power system applications. The research team reformulated the power flow equations as a combinatorial optimization problem, enabling the application of quantum algorithms to address challenges in modern electricity grids. Numerical experiments were conducted on a four-bus test system to assess both solution accuracy and computational time, providing quantitative insights into the performance of different quantum approaches. The team benchmarked results from the Quantum Approximate Optimization Algorithm, a gate-based method, against those obtained using D-Wave’s Advantage system and Fujitsu’s latest generation Digital Annealer, known as Quantum-Inspired Integrated Optimization software.
Measurements demonstrate the potential of quantum algorithms to address the computational challenges associated with modern electricity networks. The research successfully implemented and compared the performance of both gate-based and adiabatic approaches on a representative test system, revealing the trade-offs between the two paradigms and highlighting the strengths and weaknesses of each approach for solving this critical problem in electrical engineering. This work provides a crucial foundation for evaluating the viability of both gate-based and adiabatic quantum computing paradigms for power flow analysis. The combinatorial reformulation of power flow equations introduces complexity, but allows for the application of quantum optimization solvers. This breakthrough delivers valuable data for assessing the scalability and practical viability of quantum computing in the context of modern power grids.
Gate and Adiabatic Power Flow Solutions
This research presents a direct comparison of gate-based quantum computing and adiabatic quantum computing approaches to solving the AC power flow equations, a critical calculation for modern electricity networks. Scientists reformulated these equations as a combinatorial optimization problem, enabling their implementation on both types of quantum hardware. Experiments conducted on a standard four-bus test system demonstrate the viability of both paradigms, with both adiabatic quantum computing and gate-based methods yielding solutions consistent with classical power flow analysis. The findings reveal that, under the tested configuration, gate-based quantum computing, specifically the Quantum Approximate Optimization Algorithm, achieved faster iteration speeds, up to 20% quicker than the adiabatic methods.
However, adiabatic quantum computing currently demonstrates the ability to address larger, more complex problem sizes, having previously solved power system simulations with up to 1354 buses. The authors acknowledge that the small scale of the test system reflects current limitations in gate-based quantum computing technology. Future work will likely focus on scaling these methods to larger systems and improving the accuracy of gate-based implementations, potentially unlocking significant advancements in power grid optimization and control.
👉 More information
🗞 Performance Comparison of Gate-Based and Adiabatic Quantum Computing for Power Flow Analysis
🧠ArXiv: https://arxiv.org/abs/2510.13378