Project Details

[Return to Previous Page]

Optimization of Multi-Axle Vehicle Drive Systems Using AI

Company: Wabtec

Major(s):
Primary: CMPSC
Secondary: ME
Optional: CMPEN

Non-Disclosure Agreement: YES

Intellectual Property: YES

Modern locomotives are deployed across a wide range of operational scenarios, from single-unit configurations to multi-locomotive consists supporting long and heavily loaded trains. These systems operate under diverse conditions, including varying terrain, trip durations, speeds, grades, curves, and environmental factors. Currently, train operators manually adjust locomotive power in response to changing conditions. While effective, this manual control introduces inefficiencies due to human error and suboptimal coordination of locomotive resources. Additionally, the system's performance is influenced by complex interactions among mechanical and electrical subsystems—such as traction and vehicle controls, rectifiers, inverters, engines, alternators, switches, and gear assemblies. Each locomotive is equipped with multiple traction motors (typically 4, 6, or 8), each responsible for driving an axle through a gear mesh, thus rotating 2 axle mounted wheels and propelling the locomotive. Coordinated control of these axles is essential for achieving optimal propulsion and overall system efficiency. The Penn State Capstone team will be provided with representative duty cycles and system loss data. The goal is to develop AI-driven solutions that can intelligently manage and optimize drive system performance in real time. These solutions should be designed for potential integration into onboard control systems, though partial offboard implementation is acceptable if necessary.