Unmanned Aerial Vehicle trajectory tracking using Type-2 Fuzzy Logic.

摘要

The use of Unmanned Aerial Vehicles (UAVs) has expanded considerably over the last decade, especially since they have shown their value in military and civilian operations. One major problem in UAV development however is their ability to operate in an environment that contains a significant level of uncertainties. For a controller, uncertainties can take the form of plant parameter variations and disturbances. In order to investigate UAV controller design, and their performance in the presence of uncertainties, two controllers were developed using Fuzzy Logic (FL). The overall design of both controllers was identical except for the roll controller, for which one used Type-1 Fuzzy Logic (T1 FL) and the other used Type-2 Fuzzy Logic (T2 FL). The performance of both controllers was tested on a six degrees of freedom model of the Aerosonde aircraft constructed with Matlab/Simulink Aerosim Blockset. A Genetic Algorithm (GA) was developed and used to optimize the scaling gains as well as the Footprint of Uncertainty (FOU) for the T2 FL controller. The GA proved successful at optimizing both scaling gains and the FOU. Various configurations of the FOU with different spreads, sigma, were trialed and both controllers were compared in simulations with and without uncertainties. Several T2 FL controller FOU configurations outperformed the T1 FL controller showing the superiority of T2 FL. The performances of both roll controllers however, deteriorated greatly when tested in the presence of uncertainties. An adaptation mechanism, the Fuzzy Model Reference Self Tuning Controller (FMRSTC), was developed and implemented on the highest performing T2 FL controller. The FMRSTC proved to be very effective at following the model reference; however, it was unstable in the presence of disturbances. A lateral track control strategy using potential fields was also developed and tested in simulations. The lateral track control strategy proved to be very effective at reducing the cross track error between waypoints even in the presence of strong winds.;Keywords: Unmanned Aerial Vehicle, Type-2 Fuzzy Logic Controller, Genetic Algorithm Tuning, Trajectory Tracking, Lateral Tracking Control