Adaptive Safe and Energy-Efficient Driving Control of Unmanned Rail Vehicles

In the field of developing unmanned vehicles of almost all types, in the area of providing the driving control system with measurement information, most attention is nowadays paid to the development of specialized computer vision systems. That notwithstanding, in this work, a broader view of the problem of removing the driver from the control loop is proposed. Indeed, in addition to visual information, he has following data sources at his disposal: sensation of shocks, vibrations and oscillations of the car's body; sensing accelerations during speedup, slowdown and passage of curved sections; hearing noises, creaks, wheels rattling, etc. Using the entirety of the information available to the traditional human driver requires a deep study of interactions in the “moving object - motion environment” system (including the influence of the changing load and the displacement of the center of mass); selecting appropriate sensor hardware; and accommodating for these interactions in the algorithmic support.