Maximilian Posner

Abstract

Pitting damage is a critical fatigue issue on gear flanks but its progression in quenched and tempered (QT) gears, used in heavy duty use, is poorly understood. Compared to case-hardened gears with typically localized macroscopic pitting formation, QT gears show a more continuous, distributed degradation. This study develops a robust framework combining non-parametric bootstrapping with stochastic degradation models to understand and describe not only the time of failure but also the degradation including the pitting growth. An experimental program with 40 QT gear pairs and 80 test runs under varied loads is planned yielding over 300 measurements. Bootstrap-enhanced Stahl’s method captures uncertainty, while General Path and Stochastic Process Models describe deterministic growth and stochastic variability, enabling reliability prediction, lifetime estimation, and application of operating strategies during ongoing operation.

Abstract

This paper explores the application of Deep Reinforcement Learning (DRL) for optimizing Automatic Train Operation (ATO). The study aims to enhance train operations by meeting key objectives such as energy efficiency, punctuality, safety, passenger comfort, and parking accuracy. A simulation environment incorporating realistic train dynamics and track conditions was developed, utilizing discrete action and observation spaces. Various DRL algorithms, including QRDQN and PPO, were implemented and tuned using Optuna, a hyperparameter optimization framework. Results demonstrate QRDQN's superior performance across multiple evaluation journeys, highlighting its potential for achieving robust, efficient, and autonomous train operation.