A proposed methodology for the control of a semi-robotic convoy

Abstract

The purpose of this thesis is to develop a generic control law for unmanned-trail vehicles as they follow a manned lead vehicle. The development of this semi-robotic convoy control law begins with a model of an individual vehicle. Two methods are then explored of coupling these into a model of the column. A relationship between these two methods is derived. The model is then expanded to n vehicles. Utilizing a digital simulation, a three-vehicle convoy is controlled in one degree-of-freedom (DOF) using pole-placement, state-feedback control theory. The analysis shows this to be an unacceptable method of control due to the steady-state error. The 1 DOF model is then controlled with series compensation. Simulations verify that the steady-state error is eliminated. The system is then expanded into a 2 DOF system. Using the same series compensator, a 2 DOF simulation is developed. It is shown that the only additional requirement of the 2 DOF system is that the trail vehicles need to determine their orientation. This is accomplished by first saving the position and velocity profile of the lead vehicle and then developing a search algorithm to find the appropriate information. The simulation verifies that the convoy is controlled within the specifications of the system.