arXiv subjects

Yi Guo

Publications and source records attributed to Yi Guo.

2 recordsLinked to original sources

arXiv2026

Prosumer Synchronisation Risk: Impacts of Time-Varying Tariffs on Distribution Network Expansion

The rapid deployment of distributed energy resources, including heat pumps, electric vehicles, photovoltaics, and battery storage, is reshaping the operation of low-voltage networks. While distribution system operators often aim to develop time-varying tariffs to incentivise network-friendly behaviour and defer reinforcements, they risk triggering prosumer synchronisation. As prosumers follow price signals, their behaviour may synchronise, avoiding existing load peaks while creating undesirable peaks at other times. This paper quantifies the impact on low-voltage network reinforcement needs when shifting from flat tariffs to two time-varying tariffs: Time-of-Use and Real-Time Pricing. Applying a mixed-integer linear model across 471 low-voltage reference networks in Switzerland, we evaluate expansion requirements under projections of distributed energy resource deployment for 2050. In the model, prosumers operate their distributed energy resources in accordance with their subscribed electricity tariff to minimise costs, and the distribution system operator subsequently optimises reinforcements. Our results demonstrate that while adopting time-varying tariffs can reduce reinforcement investment costs by more than 50%, high adoption rates can reverse this declining trend. Specifically, once prosumer adoption exceeds the optimal thresholds of 50% for Time-of-Use and 40% for Real-Time Pricing, reinforcement needs increase by 39% and 134%, respectively, for an additional 20% of tariff adoption.

eess.SY↗

arXiv2026

CARD: Calibration via Agreement in Reverse Diffusion for Out-of-Domain MRI Segmentation

Probability calibration aligns model confidence with predictive accuracy, enabling clinicians to identify unreliable segmentation regions. This alignment breaks down under domain shift, where artifacts and unseen protocols produce confident errors. Existing post-hoc methods adapt the correction at test time, conditioning on predictive entropy, the logit pattern, or augmentation response, but each proxy is read from the terminal prediction, the very quantity that shift corrupts. This motivates reliability evidence beyond the terminal prediction, which categorical diffusion provides in two ways. First, a generative shape prior keeps a capacity-limited reference intact when appearance is corrupted, so its disagreement with the primary segmentor highlights primary-model errors. Second, every reverse step yields a class distribution, separating persistent disagreement from transient discrepancy. Aggregated over the trajectory, this disagreement correlates with Dice at 0.788, against 0.521 for a matched discriminative control. We therefore propose CARD (Calibration via Agreement in Reverse Diffusion), which maps the temporal aggregate of this disagreement to a temperature field applied per pixel across all classes, so that confidence changes while the segmentation does not. Across cardiac, prostate and brain MRI shifts, CARD lowers calibration error in 45 of 49 comparisons against the strongest baseline in each setting.

cs.CV↗