# The Impact of Feature Causality on Normal Behaviour Models for SCADA-based Wind Turbine Fault Detection

## TL;DR

This paper introduces a new taxonomy based on feature causality to evaluate how different input features influence the performance of normal behavior models in SCADA-based wind turbine fault detection.

## Contribution

It proposes a causal relation taxonomy and a framework to assess the impact of feature configurations on fault detection accuracy.

## Key findings

- Causal relations significantly affect fault detection performance.
- Different feature configurations lead to varying model accuracy.
- The framework enables systematic evaluation of feature impacts.

## Abstract

The cost of wind energy can be reduced by using SCADA data to detect faults in wind turbine components. Normal behavior models are one of the main fault detection approaches, but there is a lack of consensus in how different input features affect the results. In this work, a new taxonomy based on the causal relations between the input features and the target is presented. Based on this taxonomy, the impact of different input feature configurations on the modelling and fault detection performance is evaluated. To this end, a framework that formulates the detection of faults as a classification problem is also presented.