Toward Value-Oriented Renewable Energy Forecasting: An Iterative Learning Approach

Abstract

Energy forecasting is an essential task in power system operations. Operators usually issue forecasts and use them to schedule energy dispatch in advance. However, forecasting models are typically developed in a way that overlooks the decision value of forecasts. To bridge the gap, we design a value-oriented point forecasting approach for sequential energy dispatch problems with renewable energy sources. At the training phase, we align the training objective with the decision value, i.e., minimizing the overall operating cost. The forecasting model parameter estimation is formulated as a bilevel program. Under mild assumptions, we convert the upper-level objective into an equivalent form using the dual solutions obtained from the lower-level operation problems. In addition, a novel iterative solution strategy is proposed for the newly formulated bilevel program. Under such an iterative scheme, we show that the upper-level objective is locally linear with respect to the forecasting model output and can act as the loss function. Numerical experiments demonstrate that, compared to commonly used forecasts predicting expected realization, forecasts obtained by the proposed approach result in lower operating costs. Meanwhile, the proposed approach achieves performance comparable to that of two-stage stochastic programs, but is more computationally efficient.