Accurate Data-Based State Estimation from Power Loads Inference in Electric Power Grids

Abstract

Accurate state estimation is a crucial requirement for the reliable operation and control of electric power systems. Here, we construct a data-driven, numerical method to infer missing power load values in large-scale power grids. Given partial observations of power demands, the method estimates the operational state using a linear regression algorithm, exploiting statistical correlations within synthetic training datasets. We evaluate the performance of the method on three synthetic transmission grid test systems. Numerical experiments demonstrate the high accuracy achieved by the method in reconstructing missing demand values under various operating conditions. We further apply the method to real data for the transmission power grid of Switzerland. Despite the restricted number of observations in this dataset, the method infers missing power loads rather accurately. Furthermore, Newton-Raphson power flow solutions show that deviations between true and inferred values for power loads result in smaller deviations between true and inferred values for flows on power lines. This ensures that the estimated operational state correctly captures potential line contingencies. Overall, our results indicate that simple data-based regression techniques can provide an efficient and reliable alternative for state estimation in modern power grids.

Figures (10)

• Figure 1: Model performance for the synthetic Swiss power grid. Panels a)-e) are predictions for the largest five loads in descending order.
• Figure 2: Model performance for the synthetic Spanish power grid. Panels a)-e) are predictions for the largest five loads in descending order.
• Figure 3: Model performance for the synthetic German power grid. Panels a)-e) are predictions for the largest five loads in descending order.
• Figure 4: Model performance for inferring the loads with largest active power demand for the synthetic a) Swiss, b) Spanish, and c) German grids, as a function of the number $M$ of left-out loads.
• Figure 5: Model performance for different training dataset sizes for the synthetic a) Swiss, b) Spanish, and c) German grids for different values of $M$. The gray band for $M=10$ corresponds to the minimum and maximum prediction accuracy.