Data Requirements and Prediction Scaling for Long-Term Failure Forecasts in Wind Turbines

TL;DR

This paper analyzes factors influencing long-term failure prediction in wind turbines, highlighting dataset size as crucial and establishing a linear scaling relationship between dataset size and forecast horizon.

Contribution

It introduces a quantitative relationship between dataset size and forecast horizon, clarifying the definitions of 'big' and 'long-term' in wind turbine failure forecasting.

Findings

Dataset size is the main factor affecting forecast horizon.

An approximate linear scaling: forecast days are twice the dataset size in turbine years.

'Big' data corresponds to ten turbine years; 'long-term' forecasts span two weeks.

Abstract

We investigate the key factors that enable early failure forecasting in wind turbines. For this purpose, we analyze studies with long-term forecasts and compare their main features: prediction time, methods, targeted components, dataset size, and check the effect of using additional sensors. We found that the size of the dataset is the main factor and that an approximate linear scaling holds: the number of forecast days is twice the size of the dataset, measured in turbine years. We also observe that the data allow us to quantify the meaning of "big" and "long" in the terms "big data" and "long-term" forecasts, which are found to be ten turbine years and two weeks.