Predict+Optimize Problem in Renewable Energy Scheduling

TL;DR

The paper presents a benchmark from the IEEE-CIS Predict+Optimize Challenge for Renewable Energy Scheduling, integrating forecasting and optimization on a Monash campus microgrid. It demonstrates that stochastic and robust optimization, paired with tree-based forecasting (e.g., LightGBM ensembles), yields tangible energy-cost savings and highlights that forecast accuracy does not always predict downstream optimization performance. The open dataset, problem formulation, and competition results establish a state-of-the-art reference for decision-focused learning in energy systems, with findings that guide future research toward uncertainty-aware, scalable Predict+Optimize methods. Overall, the study confirms the practical value of integrating forecasting with optimization while outlining limitations and directions for broader applicability and multi-objective extensions.

Abstract

Predict+Optimize frameworks integrate forecasting and optimization to address real-world challenges such as renewable energy scheduling, where variability and uncertainty are critical factors. This paper benchmarks solutions from the IEEE-CIS Technical Challenge on Predict+Optimize for Renewable Energy Scheduling, focusing on forecasting renewable production and demand and optimizing energy cost. The competition attracted 49 participants in total. The top-ranked method employed stochastic optimization using LightGBM ensembles, and achieved at least a 2% reduction in energy costs compared to deterministic approaches, demonstrating that the most accurate point forecast does not necessarily guarantee the best performance in downstream optimization. The published data and problem setting establish a benchmark for further research into integrated forecasting-optimization methods for energy systems, highlighting the importance of considering forecast uncertainty in optimization models to achieve cost-effective and reliable energy management. The novelty of this work lies in its comprehensive evaluation of Predict+Optimize methodologies applied to a real-world renewable energy scheduling problem, providing insights into the scalability, generalizability, and effectiveness of the proposed solutions. Potential applications extend beyond energy systems to any domain requiring integrated forecasting and optimization, such as supply chain management, transportation planning, and financial portfolio optimization.

Figures (7)

• Figure 1: Implied data flow in the problem setting.
• Figure 2: Competition setup.
• Figure 3: Timeline of Melbourne lockdown measures in 2020 due to the COVID-19 pandemic.
• Figure 4: (a) Weather data of temperature and surface radiation from ERA5 for the location of interest. (b) Time series of electricity price for October and November 2020 (validation and test period).
• Figure 5: Input series of building load and solar power production from the Monash Clayton campus. All values are in $kW$. Building 0 and Building 3 have large outliers that have been capped at 2000. The dashed lines indicate the start of the Phase 1 test data in the competition, the solid lines indicate the start of the Phase 2 test data (i.e., data for October and November 2020, respectively).