A meta-learning based distribution system load forecasting model selection framework. (15th July 2021)

• Record Type: Journal Article
• Authors: Li, Yiyan; Zhang, Si; Hu, Rongxing; Lu, Ning
• Is Part Of: Applied energy. Volume 294(2021)
• Publication Date: 2021-07-15
• Subjects: Distribution system -- Load forecasting -- Machine learning -- Meta-learning -- Model selection -- Ensemble learning; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2021.116991
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: We introduce meta-learning concept to achieve distribution system load forecasting model selection. We make the meta-learning framework rigorously formulated, applicable and extendable. We propose a scoring-voting mechanism to improve the model selection accuracy. We create comprehensive test cases to validate the proposed methodology. Abstract: This paper presents a meta-learning based, automatic distribution system load forecasting model selection framework. The framework includes the following processes: feature extraction, candidate model preparation and labeling, offline training, and online model recommendation. Using load forecasting needs and data characteristics as input features, multiple metalearners are used to rank the candidate load forecast models based on their forecasting accuracy. Then, a scoring-voting mechanism is proposed to weights recommendations from each meta-leaner and make the final recommendations. Heterogeneous load forecasting tasks with different temporal and technical requirements at different load aggregation levels are set up to train, validate, and test the performance of the proposed framework. Simulation results demonstrate that the performance of the meta-learning based approach is satisfactory in both seen and unseen forecasting tasks.