Resilient Load Forecasting under Climate Change: Adaptive Conditional Neural Processes for Few-Shot Extreme Load Forecasting

TL;DR

It is suggested that AdaCNP can effectively mitigate the combined impact of abrupt distribution shifts and scarce extreme samples, providing a more trustworthy forecasting for resilient power system operation under extreme events.

Abstract

Extreme weather can substantially change electricity consumption behavior, causing load curves to exhibit sharp spikes and pronounced volatility. If forecasts are inaccurate during those periods, power systems are more likely to face supply shortfalls or localized overloads, forcing emergency actions such as load shedding and increasing the risk of service disruptions and public-safety impacts. This problem is inherently difficult because extreme events can trigger abrupt regime shifts in load patterns, while relevant extreme samples are rare and irregular, making reliable learning and calibration challenging. We propose AdaCNP, a probabilistic forecasting model for data-scarce condition. AdaCNP learns similarity in a shared embedding space. For each target data, it evaluates how relevant each historical context segment is to the current condition and reweights the context information accordingly. This design highlights the most informative historical evidence even when extreme samples are rare. It enables few-shot adaptation to previously unseen extreme patterns. AdaCNP also produces predictive distributions for risk-aware decision-making without expensive fine-tuning on the target domain. We evaluate AdaCNP on real-world power-system load data and compare it against a range of representative baselines. The results show that AdaCNP is more robust during extreme periods, reducing the mean squared error by 22\% relative to the strongest baseline while achieving the lowest negative log-likelihood, indicating more reliable probabilistic outputs. These findings suggest that AdaCNP can effectively mitigate the combined impact of abrupt distribution shifts and scarce extreme samples, providing a more trustworthy forecasting for resilient power system operation under extreme events.

Figures (6)

• Figure 1: (a) Standard CNP with uniform context aggregation; (b) Our proposed AdaCNP with target-aware adaptive weighting. The adaptive layer assigns relevance scores to context points based on similarity to target extreme events, amplifying critical signals while suppressing noise from irrelevant normal patterns.
• Figure 2: Regression results on a 1-D curve with phase transition (blue line) using 5, 10, and 15 context points (black stars). The color curves, as well as the surrounding shallow range, show the predicted mean and variance using AdaCNP, CNP, ANP, GP, and NP.
• Figure 3: Illustrative example of 4 extreme load curves identified in the PJM dataset. The grey curves represent normal load patterns, while the red curves are the identified extreme samples.
• Figure 4: Moving average of the training NLL loss for the load forecasting task on: (a) PJM dataset (left) and (b) ISO-NE dataset (right).
• Figure 5: Predictions of the AdaCNP, CNP, AttnCNP, GP and NP models on three extreme load curves from the test set of unseen extreme samples over a 24-hour forecasting horizon from (a) PJM dataset (left) and (b) ISO-NE dataset (right). The ground truth (blue) is the actual electricity consumption and the colored curves represent the mean predictions of each model. The shaded regions around each curve are confidence intervals, denoted by the model's prediction variance ($\mu \pm \sigma$).