Jul 2026· International Conference on Robotics and Sensor Networks· Vol 14254, pp. 142540T - 142540T-8· 0 citations· 18 references
Engineering
TL;DR
A novel probabilistic framework for net load forecasting employs Dual-coefficient network (DUCET) to model both input-space and output-space distribution characteristics, thereby enhancing robustness against distribution shifts and resolving the quantile crossing problem commonly observed in quantile regression.
Abstract
With the widespread deployment of sensor networks in modern energy systems, large-scale time-series data from distributed sources (e.g., load demand, photovoltaic generation, and wind power) provide new opportunities for intelligent system perception and predictive modeling. Accurate and uncertainty-aware net load modeling plays a critical role in supporting fault diagnosis, risk assessment, and fault-tolerant control. To overcome the deficiencies of current approaches in representing uncertainty, distribution shift, and quantile crossing, the present study develops a novel probabilistic framework for net load forecasting. It employs Dual-coefficient network (DUCET) to model both input-space and output-space distribution characteristics, thereby enhancing robustness against distribution shifts. Furthermore, a quantile loss–driven Informer architecture is adopted for modeling temporal dependencies. It can also generate probabilistic forecasts through conditional quantiles. To resolve the quantile crossing problem commonly observed in quantile regression, a quantile reconstruction strategy (QRS) is introduced, which reconstructs a coherent predictive distribution using kernel density estimation and Gaussian approximation. Experimental results on real-world net load data from Austria show that the proposed probabilistic framework surpasses benchmark methods. Additional ablation experiments confirm the validity of DUCET and QRS in improving forecasting accuracy and reliability.
This work develops a unified one-day-ahead probabilistic forecasting framework that aligns temporal resolution, reconstructs the unavailable inputs, and derives causal features, and compares a modular post-hoc residual-quantile scheme with an integrated in-model quantile-learning scheme.
S. Al-Shareeda, Gulcihan Ozdemir, H. Jeon· Electric power systems resea...· 0 citations
An integrated probabilistic forecasting framework with three linked stages: extreme-weather load identification, TimeGAN-based sample augmentation, and conformal quantile forecasting, which improves forecasting accuracy under extreme-weather conditions.
Hao Zhang, Xiyang Liu, Ruotian Gao et al.· International journal of pat...· 0 citations
Modern power transmission grids are increasingly operated under volatile load, variable generation, and near-limit line-flow conditions. In such environments, deterministic line-flow forecasting is insufficient for operational decision support because operators require calibrated risk probabilities, uncertainty intervals, and reliable early warning signals. This article proposes B-TGPRF, a Bayesian Temporal Graph Probabilistic Risk Forecaster for calibrated overload-risk forecasting in power transmission grids. The proposed framework is positioned as a hybrid probabilistic graph-temporal forecasting system rather than as a new end-to-end graph neural network; its novelty lies in the leakage-controlled sequential combination of temporal forecasting, electrical graph descriptors, calibrated Bayesian risk estimation, residual correction, and compact interval uncertainty assessment. The model integrates temporal line-flow, load, and generation features with graph-topological descriptors, operating-regime indicators, residual correction, conformal interval estimation, probability calibration, and a Bayesian risk layer. A leakage-controlled data preparation pipeline was built using an open large-scale benchmark for machine learning applications in transmission grids. The final modeling dataset contains more than 3.48 million observations, 100 selected critical lines, train-only risk thresholds, and chronological train, validation, test, and external-like scenario splits. B-TGPRF was compared with persistence baselines, linear models, Bayesian baselines, tree ensembles, boosting models, neural temporal models, and compact state-of-the-art-style temporal and graph-temporal architectures. On the strict external-like test, B-TGPRF achieved MAE = 0.3650, RMSE = 0.5478, R2 = 0.9962, Brier Score = 0.0147, and ECE = 0.0064. The results show that the proposed model provides a strong overall balance between line-flow forecasting accuracy, calibrated risk estimation, compact interval prediction, and low false-positive risk-signaling, while boosting models remain highly competitive for pure risk-class detection.
Assem Shayakhmetova, N. Tasbolatuly, Guldana Taganova et al.· Algorithms· 0 citations
High load variability and the low quality of building monitoring data pose substantial operational challenges for modern energy management systems. This study develops a robust and computationally efficient probabilistic forecasting framework by integrating data-issue handling with uncertainty calibration. Using an experimental design on a high-resolution multi-energy dataset (2018–2023), the study compares Gradient Boosting Decision Trees (GBDTs) with a linear baseline under a strict out-of-time validation protocol and Conformalized Quantile Regression (CQR). The results indicate the superiority of non-linear models: CatBoost delivers the best point-forecast accuracy, achieving a Mean Absolute Error (MAE) of 37,752.04 kW, corresponding to an 11–12% performance improvement over ElasticNet. Conformal calibration substantially improves the validity of prediction intervals, increasing the Prediction Interval Coverage Probability (PICP) from 82.22% to 87.28%, thereby approaching the nominal 90% confidence target without imposing strong distributional assumptions. Further ablation analyses reveal that rolling-window features contribute more to accuracy than external weather variables. Overall, these findings provide a practical contribution in the form of a forecasting method that is not only accurate but also statistically reliable in estimating operational risk, thereby bridging the gap between industry demands for robust systems and the constraints imposed by real-world data quality.
Lasmedi Afuan, Agus Darmawan, Raden Demas Amirul Plawirakusumah et al.· Engineering, Technology &...· 0 citations
With the large-scale integration of distributed photovoltaic generation (PV) into modern distribution networks, the inherent stochasticity and volatility of renewable energy outputs have imposed non-negligible impacts on the secure and economic operation of power systems. Conventional probabilistic power flow (PPF) methods are limited in accurately modeling source–load uncertainty and, more importantly, in capturing complex nonlinear and time-varying dependence among multiple renewable energy sources. To address these issues, this paper proposes a novel PPF calculation framework based on advanced source-load modeling and time-varying D-vine Copula. Firstly, an enhanced finite mixture Beta model and a Gaussian cluster mixture model are developed to characterize the uncertainty of PV output and load demand, respectively. Secondly, a time-varying D-vine Copula model based on the generalized autoregressive score framework is constructed. And a two-stage regularized profile likelihood estimation method is proposed to estimate correlation parameters, capturing the dynamic nonlinear dependence among multiple PV generators. Finally, the de-randomized Sobol sequence-based Quasi-Monte Carlo method is adopted to perform stochastic power flow calculation. Simulation results on a real-world 129-bus distribution system in East China verify the accuracy and effectiveness of the proposed method.
The increasing penetration of variable renewable energy (VRE) introduces profound stochastic uncertainties into power systems, rendering traditional deterministic security assessments insufficient. This paper addresses this challenge by developing a comprehensive, AI-ready operational risk assessment framework based on a data-driven Probabilistic Power Flow (PPF) approach. Unlike conventional methods that rely on assumed statistical distributions, the proposed methodology employs a statistical learning pipeline to identify suitable probability density functions directly from historical SCADA telemetry, thereby ensuring high-fidelity input modeling. Furthermore, a cluster-based correlation strategy, rooted in unsupervised pattern recognition, is constructed to rigorously capture the spatiotemporal dependencies among renewable generation sources. The probabilistic framework is executed via an enhanced Monte Carlo simulation (MCS) utilizing Latin Hypercube Sampling (LHS) to improve computational efficiency. Validated through a case study on the Central Vietnam Power System, the results demonstrate that the proposed framework effectively uncovers hidden operational vulnerabilities, specifically identifying a 4.57% probability of exceeding the 90% branch-loading warning threshold on the critical 220 kV Quy Nhon - Tuy Hoa transmission line, while the probability of actual thermal overload above 100% is nearly zero. This near-limit operating risk profile may be masked by conventional deterministic snapshots. The findings emphasize that accurately modeling the correlation structure of renewable sources is indispensable for realistic risk quantification, providing system operators with a robust, algorithm-driven decision-support tool for grid security management.
L. Nhan, Lê Thị Phương Thảo, L. H. Lam et al.· E3S Web of Conferences· 0 citations