A Machine Learning Approach to Aircraft Sensor Error Detection and Correction

Abstract

Sensors are crucial to modern mechanical systems. The location of these sensors can often make them vulnerable to outside interferences and failures, and the use of sensors over a lifetime can cause degradation and lead to failure. If a system has access to redundant sensor output, it can be trained to autonomously recognize errors in faulty sensors and learn to correct them. In this work, we develop a novel data-driven approach to detect sensor failures and predict the corrected sensor data using machine learning methods in an offline/online paradigm. Autocorrelation is shown to provide a global feature of failure data capable of accurately classifying the state of a sensor to determine if a failure is occurring. Feature selection of the redundant sensor data in combination with k-nearest neighbors regression is used to predict the corrected sensor data rapidly, while the system is operational. We demonstrate our methodology on flight data from a four-engine commercial jet that contains failures in the pitot static system resulting in inaccurate airspeed measurements.