Multi-variable Controllers for Cooperative Flight of Multi-Fixed Wing UAVs

Abstract

This paper proposes a collaborative control system to be designed for multi-UAV. This makes it easy to perform many tasks at the same time and with high accuracy. Therefore, this cooperative control and guidance subsystems of the aircraft should have robust performance against sensors noise and wind disturbances. Four types of control algorithms were designed for a single Aerosonde UAV autopilot. This is to pick up which control algorithm is the best. As such, this control algorithm is proposed to be designed for the cooperative flight control system. Two classical control algorithms and two intelligent control algorithms have been proposed for the autopilot design of a single Aerosonde UAV. The first classical controller proposed is genetically tuned PID, while the second classical controller proposed is the fractional order PID. The first intelligent controller proposed for autopilot system is the Fuzzy logic controller known as FLC, while the second intelligent controller proposed is the adaptive neuro fuzzy inference system known as ANFIS. The proposed control algorithms have been applied to the nonlinear multivariable system of Aerosonde UAV. The analysis of simulation results assure that ANFIS is the best performance and the most robust control algorithm proposed. As such, ANFIS controller has been selected to be the cooperative flight controller system either in the low-level of a single UAV and in the top-level of multi-UAVs. Sometimes, classical controllers are preferred because of their simplicity in design. If this is the case, the simulation results assure that the genetically tuned fractional order PID controller- which proposed here for the first time with UAVs- is better than genetically tuned PID.