Electro-Hydrostatic Actuator-Based Steering for Non-Road Mobile Machinery

Abstract

In the dynamic realm of Non-Road Mobile Machinery (NRMM), technological evolution is driving new standards and challenges. Steering, as one critical aspect of NRMM, has traditionally relied on hydrostatic systems renowned for their reliability and inherent safety. However, these conventional steering mechanisms, inherently inefficient, are proving insufficient for the automation and electrification trends shaping the industry. As cutting-edge technologies continue to redefine NRMM, the limitations of traditional steering systems underscore the pressing need for innovative alternatives that offer enhanced adaptability, superior energy efficiency, and safer operation. This thesis explores Electro-Hydrostatic Actuator (EHA)-based steering systems as an innovative solution, targeting three key goals: compliance with international safety standards, enhanced energy efficiency, and integration of Artificial Intelligence (AI) based technologies for advanced fault detection.

Two novel EHA-based steering topologies are proposed: a fully Steer-by-Wire (SbW) system and a hybrid system with passive redundancy. The SbW system features dual channels—an electric motor-driven EHA and a proportional valve-based backup—achieving a safety Performance Level 'd' per ISO standards with low diagnostic coverage (60–90%). The hybrid system pairs an EHA with a customized orbital valve, offering manual fallback and efficiencies ranging from 28–50%, with throttling losses below 15%. Using simulation models and experimental data from a wheel loader, comparative analyses demonstrate that EHA-based systems reach up to 81.3% efficiency—outperforming conventional systems, which lose 51–60% of energy to throttling—marking a remarkable improvement.

Safety is further enhanced through AI-based fault detection, implemented in a co-simulation model. This system identifies critical faults (e.g., power loss, uncommanded steering) with 100% accuracy in 155–337 milliseconds, enabling swift redundancy activation. By integrating safety, efficiency, and intelligence, this research fills a vital gap in NRMM steering technology, delivering a practical framework for sustainable and reliable systems in the field of NRMM. Future efforts will focus on field testing and expanding AI datasets to broaden applicability, paving the way for widespread adoption of advanced steering solutions in NRMM.