Joint Prediction of Remaining Useful Life and Failure Type of Train Wheelsets: A Multi-task Learning Approach
This work aims to improve the prediction of train wheelset failures for train operators, potentially reducing disruptions and derailments.
This paper addresses the prediction of train wheelset remaining useful life (RUL) and failure type. It proposes a multi-task learning approach that jointly predicts RUL (regression) and failure type (classification), achieving a 3% improvement in prediction accuracy over single-task learning methods.
The failures of train wheels account for disruptions of train operations and even a large portion of train derailments. Remaining useful life (RUL) of a wheelset measures the how soon the next failure will arrive, and the failure type reveals how severe the failure will be. RUL prediction is a regression task, whereas failure type is a classification task. In this paper, we propose a multi-task learning approach to jointly accomplish these two tasks by using a common input space to achieve more desirable results. We develop a convex optimization formulation to integrate both least square loss and the negative maximum likelihood of logistic regression, and model the joint sparsity as the L2/L1 norm of the model parameters to couple feature selection across tasks. The experiment results show that our method outperforms the single task learning method by 3% in prediction accuracy.