Minimum-time open-loop and closed-loop optimal guidance with GA-PSO and neural fuzzy for Samarai MAV flight

Abstract

Nonmilitary applications of unmanned air vehicles (UAVs), especially as rescue vehicles to obtain information have increased in recent years. Small UAVs (SUAV) were used in nonmilitary projects, where the focus was on autonomous vehicles; however, microaerial vehicles (MAVs), such as Samarai, make use of radio control, and they were not autonomous vehicles [1]. Autonomous UAVs need sophisticated guidance, control, and navigation systems with conventional aircraft flying processes. Most of current UAVs use ground-station commands for guidance purposes, although autonomous UAVs tend to reduce dependency on ground-station command via autonomous algorithms, such as intelligent methods. Therefore, ground-based air control may not suffice to cover every flying mission. Thus, self-contained, onboard guidance via intelligent methods, such as fuzzy logic, has been considered to achieve this goal. Lockheed Martin’s Intelligent Robotics Laboratory has spent the last 5 years developing an unmanned craft to replicate the air vehicle motion. The idea was based on the motion of maple seed, which whirls softly to the ground after dropping from maple tree. This motion, which is similar to a one-winged helicopter, inspired Lockheed Martin to design a new type of flying vehicles, beneficial in military and nonmilitary surveillance. Maple seeds disperse themselves by auto-rotating passively by using a single wing as they descend from trees, thereby ensuring they are widely scattered. Inspired by these flight concepts, Lockheed Martin Advanced Technology Laboratories developed the Samarai MAV, a 30-cm-radius maple seedlike aircraft that can take off and land vertically and fly laterally, like a helicopter, due to the intrinsic stability of the maple seeds’ nature [2] (Figure 1).