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

In this research, time-optimal open-loop and closed-loop guidance laws for MAVs, such as Samarai, were investigated. Open-loop optimal solutions were achieved with a new technique,based on using mathematical methods, such as calculus of variation and the GA-PSO optimization algorithm. This novel method could overcome difficulties of the usual optimal control methods. The TS approach developed closed-loop guidance that dealt with the novel neural-fuzzy training algorithm based on open-loop optimal trajectories to achieve closed-loop guidance. By feed forwarding the optimal control as a nominal optimum trajectory plus deviation, closed-loop optimal control can be generated. On the other hand, (TS) fuzzy logic needs storing nominal optimal trajectories to train the system. As discussed, the responses of the optimal closed-loop guidance were achieved, and the robustness of closed-loop policies against noises and/or disturbances was studied. However, it was difficult to obtain closed-loop solutions when the other methods, such as dynamic programming or analytical methods, were considered. The case mentioned is an important innovative idea for autonomous MAVs, whose results are important for further developments of Samarai monocopter. The results of this work demonstrated that the closed-loop optimal responses based on TS fuzzy logic methods can damp disturbances for minimum time criterion. However, the result of fuzzy logic is perfectly suitable, where higher noises were applied. Therefore, it can be concluded that fuzzy logic was robust and efficient in eliminating perturbations. It is proposed that aerospace engineers, such as primary trajectory designers, use TS fuzzy training guidance because its robustness was guaranteed by this work, especially for MAVs, such as Samarai.