arXiv subjects

S. Pineda

Publications and source records attributed to S. Pineda.

5 recordsLinked to original sources

arXiv2026

End-to-End Pseudo-Measurement Learning for State Estimation under Limited Observability

Distribution System State Estimation (DSSE) is becoming increasingly important with the integration of Distributed Energy Resources (DERs) and the active operation of distribution networks (DNs), but it remains challenging due to the limited and heterogeneous monitoring infrastructure available in these networks. To address this challenge, this paper proposes a novel DSSE framework that restores observability through data-driven pseudo-measurements generated by a Neural Network (NN), while preserving the exact non-linear AC power flow model within a classical Weighted Least Squares (WLS) estimator. Unlike conventional approaches that generate pseudo-measurements independently of the physical estimation process, the proposed method explicitly couples both components through an end-to-end learning formulation. Specifically, the WLS estimator is embedded as a layer within the NN architecture, enabling implicit differentiation to propagate the impact of pseudo-measurements on the final estimation error back to the NN parameters. Extensive numerical experiments on the IEEE 30-bus and IEEE 33-bus systems demonstrate that the proposed framework consistently outperforms state-of-the-art methods in state estimation (SE) accuracy under a wide range of loading conditions and measurement configurations.

math.OC↗

arXiv2025

Contextual Robust State Estimation in Distribution Systems with Real-Time Unobservability and Scarce Data

Distribution system state estimation faces a double information shortage: real-time measurements are often too sparse to guarantee observability, while historical data, despite being abundant, rapidly loses representativeness as operating conditions evolve. This paper proposes a data-driven methodology that addresses both challenges simultaneously. We first introduce a $K$-nearest neighbor estimator that approximates the conditional distribution of delayed measurements given the available real-time context, casting state estimation as a weighted least squares problem over plausible pseudo-measurement scenarios. Building on this, we develop a Contextual Robust State Estimator (CR-SE) that explicitly accounts for the statistical uncertainty arising from the estimation of pseudo-measurements. Rather than protecting against gross errors or bad data, CR-SE hedges against potential misspecification of the conditional distribution by optimizing over an adversarial reweighting of the nearest neighbors subject to contextual proximity and monotonicity constraints. The resulting min-max problem admits an equivalent single-level reformulation with essentially the same computational tractability as the baseline estimator. Furthermore, the robustness parameter is selected online through a fully data-driven validation procedure based on the most recent training instance. Numerical experiments on the IEEE 38-bus and modified IEEE 123-bus radial networks, as well as the meshed IEEE 30-bus active distribution network, show that CR-SE consistently outperforms the baseline across a wide range of measurement availability, training sizes, and loading conditions, achieving tail-error reductions of up to 18%, with the largest gains occurring precisely in the most challenging scenarios.

math.OC↗

arXiv2023

Learning-based State Estimation in Distribution Systems with Limited Real-Time Measurements

The task of state estimation in active distribution systems faces a major challenge due to the integration of different measurements with multiple reporting rates. As a result, distribution systems are essentially unobservable in real time, indicating the existence of multiple states that result in identical values for the available measurements. Certain existing approaches utilize historical data to infer the relationship between real-time available measurements and the state. Other learning-based methods aim to estimate the measurements acquired with a delay, generating pseudo-measurements. Our paper presents a methodology that utilizes the outcome of an unobservable state estimator to exploit information on the joint probability distribution between real-time available measurements and delayed ones. Through numerical simulations conducted on a realistic distribution grid with insufficient real-time measurements, the proposed procedure showcases superior performance compared to existing state forecasting approaches and those relying on inferred pseudo-measurements.

math.OC↗

arXiv2022

A Functional Data Analysis Approach to Evolution Outlier Detection for Grouped Smart Meters

Smart metering infrastructures collect data almost continuously in the form of fine-grained long time series. These massive data series often have common daily patterns that are repeated between similar days or seasons and shared among grouped meters. Within this context, we propose an unsupervised method to highlight individuals with abnormal daily dependency patterns, which we term evolution outliers. To this end, we approach the problem from the standpoint of Functional Data Analysis (FDA) and we use the concept of functional depth to exploit the dynamic group structure and isolate individual meters with a different evolution. The performance of the proposal is first evaluated empirically through a simulation exercise under different evolution scenarios. Subsequently, the importance and need for an evolution outlier detection method is shown by using actual smart-metering data corresponding to photo-voltaic energy generation and circuit voltage records. Here, our proposal detects outliers that might go unnoticed by other approaches of the literature that have demonstrated to be effective capturing magnitude and shape abnormalities.

stat.ME↗

arXiv2021

Is learning for the unit commitment problem a low-hanging fruit?

The blast wave of machine learning and artificial intelligence has also reached the power systems community, and amid the frenzy of methods and black-box tools that have been left in its wake, it is sometimes difficult to perceive a glimmer of Occam's razor principle. In this letter, we use the unit commitment problem (UCP), an NP-hard mathematical program that is fundamental to power system operations, to show that simplicity must guide any strategy to solve it, in particular those that are based on learning from past UCP instances. To this end, we apply a naive algorithm to produce candidate solutions to the UCP and show, using a variety of realistically sized power systems, that we are able to find optimal or quasi-optimal solutions with remarkable speedups. Our claim is thus that any sophistication of the learning method must be backed up with a statistically significant improvement of the results in this letter.

math.OC↗