To raise awareness of the environmental impact of deep learning (DL), numerous studies have estimated the energy consumption of DL systems. However, energy estimates during DL training often rely on unverified assumptions. This work addresses that gap by investigating how model architecture and training environment affect energy consumption. We train a variety of computer vision models and collect energy consumption and accuracy metrics to analyze their trade-offs across configurations. Our results show that selecting the right model-training environment combination can reduce training energy consumption by up to 80.68% with less than 2% loss in F1 score. We find a significant interaction effect between model and training environment: energy efficiency improves when GPU computational power scales with model complexity. Moreover, we demonstrate that common estimation practices, such as using FLOPs or GPU TDP, fail to capture these dynamics and can lead to substantial errors. To address these shortcomings, we propose the Stable Training Epoch Projection (STEP) and the Pre-training Regression-based Estimation (PRE) methods. Our evaluation demonstrates that STEP and PRE achieve reductions in Root Mean Squared Error (RMSE) up to 97% and 84%, respectively, when compared to existing estimation tools.
Transmission System Operators routinely use transmission switching as a tool to manage congestion and ensure system security. Motivated by sub-transmission operations at RTE, this paper considers the Optimal Transmission Switching with De-energization (OTSD), which captures potential loss of connectivity (and therefore localized blackout) following loss of transmission elements. While directly relevant to real-life operations, this problem has received very little attention in the literature. The paper proposes a new mixed-integer linear programming formulation for OTSD that represents post-contingency loss of connectivity without requiring additional binary variables. This new formulation provides the foundation for a fast, iterative heuristic algorithm. Computational experiments confirms that state-of-the-art optimization solvers struggle to solve the extensive formulation of OTSD, often failing to find even trivial solutions within reasonable time. In contrast, numerical results demonstrate the efficiency of the proposed heuristic, which finds high-quality feasible solutions 100-1000x faster than using Gurobi.
Reliability assessment of engineering systems often requires repeated evaluations of limit-state functions that may rely on computationally expensive high-fidelity models, rendering direct sampling-based reliability analysis impractical. An effective solution is to approximate the limit-state function with a surrogate model that can be iteratively refined through active learning, thereby reducing the number of computationally expensive evaluations. At each iteration, an acquisition strategy selects the next sample for evaluation by balancing two competing objectives: exploration, to reduce global predictive uncertainty, and exploitation, to improve accuracy near the failure boundary. Conventional strategies such as the U-function, EFF, ERF, REIF, and portfolio-based schemes encode this balance through single pointwise scores, concealing the underlying tradeoff. In this work, we formulate sample acquisition as a multi-objective optimization (MOO) problem in which exploration and exploitation are explicit competing objectives, yielding a compact Pareto set that provides a quantifiable trade-off representation. To select samples from the Pareto set, we investigate principled MOO criteria and propose adaptive trade-off rules, including a scheduled exploration-to-exploitation shift and a reliability-aware selection rule. Across diverse limit-state functions, we evaluate all tested strategies through relative failure-probability error trajectories, sample-efficiency comparisons, and global rankings, showing that the adaptive MOO-based strategies achieve robust overall performance while consistently meeting strict error targets.
Distributed energy resources (DERs) are transforming power networks, challenging traditional operational methods, and requiring new coordination mechanisms. To address this challenge, this paper introduces SecuLEx (Secure Limit Exchange), a market-based paradigm for allocating and trading power injection and withdrawal limits, known as dynamic operating envelopes (DOEs). Under this paradigm, distribution system operators (DSOs) first assign initial DOEs to customers through a fair allocation mechanism. These limits can be exchanged afterward through a market, allowing customers to reallocate them according to their needs while ensuring network operational constraints. We formalize SecuLEx and illustrate DOE allocation and market exchanges on a small-scale low-voltage (LV) network. In this case study, SecuLEx reduces renewable curtailment and improves grid utilization and social welfare compared to traditional approaches.
Effective tourism management requires understanding how attraction attributes and supporting factors shape tourists’ perceived attractiveness, and how these in turn translate into visitation. We develop a tourist-centered framework using 610,136 reviews across 2128 heritage attractions in 143 cities, measuring multi-scale perceived attractiveness with a large language model pipeline and visitation with review volume, then identifying the factors that influence both metrics using XGBoost and SHAP. Results show that (1) perceived attractiveness and visitation share a multi-core spatial structure, while visitation displays a polarized long-tail distribution; (2) attraction attributes, especially official designation, visual quality, and ticket pricing, dominate both metrics but shape them differently; (3) positively perceived social, economic, scientific, historic and aesthetical values increase visitation, whereas political value is negatively associated; and (4) accessibility and spatial agglomeration also contribute positively. Practically, multi-scale benchmarking identifies underperforming attractions, while attraction-scale value profiles and the identified influencing factors guide targeted strategies.
