DEWP: Deep Expansion Learning for Wind Power Forecasting

TL;DR

DEWP introduces a deep expansion learning framework for wind power forecasting that explicitly models complex dependencies in multivariate time series. It stacks primitive blocks—variable expansion via 1-D convolutions, time expansion with Fourier-constrained backcast/forecast, and a per-stack inference module with multi-head attention—and uses expansion residue learning to combine stack outputs. Empirical results on two real-world turbines demonstrate state-of-the-art performance across short- and long-term horizons, with ablations confirming the importance of each component. The approach offers practical benefits for grid operation and energy markets by delivering more accurate, timely wind power predictions while highlighting considerations for deployment and computation. The work thereby advances scalable, attentive, multi-view forecasting in renewable energy systems.

Abstract

Wind is one kind of high-efficient, environmentally-friendly and cost-effective energy source. Wind power, as one of the largest renewable energy in the world, has been playing a more and more important role in supplying electricity. Though growing dramatically in recent years, the amount of generated wind power can be directly or latently affected by multiple uncertain factors, such as wind speed, wind direction, temperatures, etc. More importantly, there exist very complicated dependencies of the generated power on the latent composition of these multiple time-evolving variables, which are always ignored by existing works and thus largely hinder the prediction performances. To this end, we propose DEWP, a novel Deep Expansion learning for Wind Power forecasting framework to carefully model the complicated dependencies with adequate expressiveness. DEWP starts with a stack-by-stack architecture, where each stack is composed of (i) a variable expansion block that makes use of convolutional layers to capture dependencies among multiple variables; (ii) a time expansion block that applies Fourier series and backcast/forecast mechanism to learn temporal dependencies in sequential patterns. These two tailored blocks expand raw inputs into different latent feature spaces which can model different levels of dependencies of time-evolving sequential data. Moreover, we propose an inference block corresponding for each stack, which applies multi-head self-attentions to acquire attentive features and maps expanded latent representations into generated wind power. In addition, to make DEWP more expressive in handling deep neural architectures, we adapt doubly residue learning to process stack-by-stack outputs. Finally, we present extensive experiments in the real-world wind power forecasting application on two datasets from two different turbines to demonstrate the effectiveness of our approach.

Figures (10)

• Figure 1: Turbine R80736 and Turbine R80721 at 'La-Haute-Borne' wind farm.
• Figure 2: An example of several related sequential factor variables and target (generated power) of 6 months since Janurary 1st, 2013 from turbine R80736.
• Figure 3: The distribution of wind speed, wind direction, and generated power of a year since Janurary 1st, 2013 from turbine R80736.
• Figure 4: Framework Overview. The left part shows the input and output of the framework; the middle part shows the main architecture of DEWP; the right part shows inner components of each stack.
• Figure 5: Ablation tests of Short-Term with different experimental setups on R80736 and R80721 datasets.