Graph-Based Reinforcement Learning for Power Distribution Voltage Regulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Power distribution systems face challenges in voltage regulation and reactive power flow due to radial topology, leading to undervoltage issues at distant nodes, and existing optimization methods struggle to efficiently manage controllable grid assets across large systems.
Innovation Solution
The implementation of graph-based reinforcement learning to control voltage and reactive power flow by training a control policy using graph neural networks, which processes system state observations and topological information to optimize the status of controllable assets like voltage regulators, capacitors, and batteries, thereby regulating voltage and minimizing power losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If radial topology is used in power distribution systems, then system simplicity and ease of operation are improved, but voltage regulation deteriorates at distant nodes
Solution Approach 1:
The patent changes the control parameters of controllable grid assets (voltage regulators, capacitors, batteries) to dynamically adjust voltage and reactive power flow. By modifying operational parameters rather than system topology, the radial structure is preserved while voltage regulation is improved through optimized asset dispatch.
2Reliability
If traditional optimization methods are used for volt-var control, then local voltage regulation is improved, but scalability to large systems deteriorates
Solution Approach 1:
The patent replaces traditional mechanical optimization methods with graph-based reinforcement learning. This substitution enables the system to handle large-scale power distribution networks efficiently by using neural networks to learn optimal control policies from graph representations, achieving both good voltage regulation and system scalability.
3Reliability
If more controllable grid assets are deployed to improve voltage regulation, then voltage profile quality is improved, but system complexity and cost increase
Solution Approach 1:
The patent enables controllable grid assets to self-optimize their operation through reinforcement learning. The graph-based RL model automatically determines optimal dispatch strategies for voltage regulators, capacitors, and batteries, reducing the need for complex centralized control systems while maintaining good voltage profile quality.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for controlling a power distribution system having a number of nodes and controllable grid assets associated with at least some of the node includes acquiring observations via measurement signals associated with respective nodes and generating a graph representation of a system state based on the observations and topological information of the power distribution system. The topological information is used to determine edges defining connections between nodes. The observations are used to determine nodal features of respective nodes, which are indicative of a measured electrical quantity and a status of controllable grid assets associated with the respective node. The graph representation is processed using a reinforcement learned control policy to output a control action for effecting a change of status of one or more of the controllable grid assets, to regulate voltage and reactive power flow in the power distribution system based on a volt-var optimization objective.