Using Fuzzy Inference Systems for blending humans' and automatic control on combat ground vehicles.

Hebert Azevedo Sá¹; Armando Morado Ferreira²; Mauro Speranza Neto¹; Marco Antonio Meggiolaro¹

¹ PUC-Rio; ² ECEME

doi:10.20906/CPS/COB-2015-1164

Resumo

Human operators' and robots' shared control techniques are encouraging the research and development of Advanced Driver Assistance Systems (ADAS) and semi-autonomous control for automotive applications. In semi-autonomous control, the main feature is the blending of control signals generated by an automatic controller with the control from a human operator. In recent research, various techniques have been proposed to achieve this mixture of control signals. However, the use of Artificial Intelligence for that purpose is still scarce. The purpose of this work is to use Fuzzy Inference Systems to design an appropriate semi-autonomous control system for military vehicles, from which the choice of the automatic controller intervention level would be achieved. The results from simulations of the military vehicle are presented, while operated in dangerous situations, where the intervention is made necessary, e.g. in the presence of hostile enemy threats or in highly destabilizing maneuvers. The behavior of the controller's intervention variable is presented through its evolution curves and indicates its increase accordingly to the growth of the threat level for which the vehicle is exposed to. The results are critically analyzed, and it is concluded that the use of the proposed system will result in a qualitative increase in vehicle's safety, making it a more efficient military system, with greater operational capacity and enhancing the skills of its driver.

Palavras-chave: Fuzzy Systems; Semi-autonomous Control; Military Vehicles