Self-constructing controller for autonomous vehicle

• Main Author: Muhammad Amirul Afiq
• Other Authors: Mahardhika Pratama
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: https://hdl.handle.net/10356/141682
• Institution: Nanyang Technological University

The Unmanned Aerial Vehicle (UAV) is one of the marvels of technologies that has constantly evolved for many more applications. The UAV, which was once used only for military applications as a means of a weapon system through air strikes in World War II, has now expanded beyond that. This is thanks to advancement in technology, which has pathed ways for UAVs (drones) to be accessible for the masses. Various modern techniques have been developed and implemented in modern UAVs such that it has partial autonomous capabilities. One of such methods is the ability to operate itself to fly along a predetermined path. While such research has its merits in building potentially fully autonomous UAV, the common challenge has always been to provide a stable and consistent controller for it. This is due to the complexity of the UAV hardware where it is difficult and time consuming to tune and optimize to operate perfectly under various conditions. This project explores the performance of a Generic Evolving Self-Organizing Neuro-Fuzzy Control of Bio-inspired Unmanned Aerial Vehicles (G-Controller) when implemented on an autonomous UAV. The G-Controller comprises the combination of Sliding Mode Control (SMC) with generic evolving neuro-fuzzy system (GENEFIS). The goal of the G-controller is to develop a self-organizing and evolving controller that’s capable of handling uncertainties in the environment. Users are also not required to be domain experts and no offline training is required. Through testing and simulating the use of this controller on an autonomous UAV plant, we can determine the viability of implementing the UAV for realworld applications where environmental conditions may vary.