Nobody but You: Sensor Selection for Voltage Regulation in Smart Grid

Rukun Mao, Husheng LiarXiv 2011 · arXiv

Introduction

Modern smart grids increasingly rely on Distributed Energy Resources (DERs), such as solar panels and wind turbines, to maintain stable voltage levels. Effective regulation requires a central controller to receive real-time data from various sensors distributed across the grid. However, these sensors often compete for limited bandwidth on a single wireless communication channel. Traditional "round-robin" polling, where sensors take fixed turns, often fails to prioritize critical data from volatile areas of the grid.

This paper addresses the challenge of optimal sensor selection: determining which sensor should report in each time slot to minimize voltage fluctuations and control costs. The authors propose a "state-aware" strategy that models the grid and its communication network as a hybrid dynamical system, achieving a significant performance improvement over standard scheduling methods.

Methodology

The problem is framed using Hybrid Dynamical Systems (HDS) theory, where the continuous state represents grid voltage and the discrete "switch" represents sensor activation. The framework relies on three core components:

  1. State Estimation: A Kalman Filter estimates the full grid state from partial, noisy data provided by the active sensor.
  2. Feedback Control: A Linear Quadratic Regulator (LQR) determines optimal DER adjustments to minimize voltage deviation and control effort.
  3. Optimization: A Sliding Window Algorithm is employed to manage computational complexity. By looking ahead a few steps to find a local optimal sequence, the system can make real-time decisions without the exponential cost of long-term planning.

Experimental Setup

The strategy was evaluated using a simulated 3-bus power system model. This setup included three voltage-regulating DERs and three monitoring sensors. The simulation incorporated mathematical matrices representing the system's physical properties, process noise, and measurement noise to test the algorithm's robustness.

Key Results and Insights

The proposed state-aware strategy demonstrates a clear advantage over traditional methods. By making the Media Access Control (MAC) layer aware of the physical state of the grid, the system becomes more resilient and efficient.

Comparison of unit cost over time with and without the sensor selection strategy.

Key findings include:

  • 40% Performance Improvement: The strategy reduced total system cost—a metric combining voltage deviation and control effort—by approximately 40% compared to round-robin scheduling.
  • Faster Convergence: As shown in the voltage derivation plots, the proposed strategy eliminates deviations by time step $k=30$, whereas significant fluctuations persist under the baseline scheme beyond $k=40$.
  • Noise Compensation: The algorithm intelligently prioritizes sensors in unstable or high-noise areas. In simulations, a sensor with higher measurement noise was automatically granted more communication slots (45%) to ensure data accuracy.
Voltage derivation comparison showing faster convergence and smoother regulation with the proposed strategy.

Conclusion and Future Work

While the sliding window approach significantly improves grid stability, several areas remain for future exploration. Current models assume a perfect communication channel; real-world factors like transmission delays and interference must be integrated. Additionally, moving from a centralized controller to a decentralized "microgrid" architecture where sensors coordinate autonomously represents a key next step in scaling this technology for massive, complex power grids.

Summary updated

Abstract

The increasing availability of distributed energy resources (DERs) and sensors in smart grid, as well as overlaying communication network, provides substantial potential benefits for improving the power system's reliability. In this paper, the problem of sensor selection is studied for the MAC layer design of wireless sensor networks for regulating the voltages in smart grid. The framework of hybrid dynamical system is proposed, using Kalman filter for voltage state estimation and LQR feedback control for voltage adjustment. The approach to obtain the optimal sensor selection sequence is studied. A sub- optimal sequence is obtained by applying the sliding window algorithm. Simulation results show that the proposed sensor selection strategy achieves a 40% performance gain over the baseline algorithm of the round-robin sensor polling.
BibTeX
@misc{mao2011nobody,
  author = {Rukun Mao and Husheng Li},
  title = {{Nobody but You: Sensor Selection for Voltage Regulation in Smart Grid}},
  year = {2011},
  eprint = {1103.5441},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1103.5441},
}
Figure