Transmission Interface Power Flow Adjustment: A Deep Reinforcement Learning Approach based on Multi-task Attribution Map

TL;DR

This paper tackles the coupling and uncertainty in transmission interface power flow adjustment by formulating a multi-task reinforcement learning problem and introducing a Multi-Task Attribution Map (MAM). The approach combines dual Graph Convolution Networks for node features, a task-representation encoder to generate task-specific queries, and a dueling Deep Q Network to output near-optimal generation-dispatch actions, with an attribution mechanism that yields a compact, task-adaptive state. Empirical results on IEEE 118-bus, a 300-bus Chinese system, and a 9241-bus European system show that MAM outperforms state-of-the-art DRL baselines and traditional OPF in both single-interface and multi-interface settings, while delivering significant improvements in inference speed and interpretability. The work demonstrates the practicality of learning multiple transmission-interface adjustment tasks jointly and highlights the potential for improved security and economic operation in large-scale power systems.

Abstract

Transmission interface power flow adjustment is a critical measure to ensure the security and economy operation of power systems. However, conventional model-based adjustment schemes are limited by the increasing variations and uncertainties occur in power systems, where the adjustment problems of different transmission interfaces are often treated as several independent tasks, ignoring their coupling relationship and even leading to conflict decisions. In this paper, we introduce a novel data-driven deep reinforcement learning (DRL) approach, to handle multiple power flow adjustment tasks jointly instead of learning each task from scratch. At the heart of the proposed method is a multi-task attribution map (MAM), which enables the DRL agent to explicitly attribute each transmission interface task to different power system nodes with task-adaptive attention weights. Based on this MAM, the agent can further provide effective strategies to solve the multi-task adjustment problem with a near-optimal operation cost. Simulation results on the IEEE 118-bus system, a realistic 300-bus system in China, and a very large European system with 9241 buses demonstrate that the proposed method significantly improves the performance compared with several baseline methods, and exhibits high interpretability with the learnable MAM.

Figures (8)

• Figure 1: Illustration of the proposed method.
• Figure 2: Illustration of the IEEE 118-bus system. Different transmission interfaces are represented by different colors. Please zoom for better view.
• Figure 3: Illustration of the realistic 300-bus system in China. Different transmission interfaces are represented by different colors.
• Figure 4: Illustration of the European 9241-bus system from the PEGASE project. Different transmission interfaces are represented by different colors.
• Figure 5: Learning curves of our method and baselines in both single-interface tasks and multi-interface tasks under the multi-task setting. All experimental results are illustrated with the median performance and one standard deviation (shaded region) over 5 random seeds for a fair comparison.