Design of a fuzzy GS-PID controller for payload drops of varying mass for a quadrotor

• Main Author: Gue, Ivan Henderson V.
• Format: text
• Language: English
• Published: Animo Repository 2014
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/4756
• Institution: De La Salle University

Unmanned Aerial Vehicles (UAVs) have become a popular research topic among researchers for the past few years. It is because of their potential on a wide variety of applications that intrigues these researchers to pursue this topic. One of the most popular UAV is the quadrotor type. With VTOL capability and simple mechanics, it is no wonder that this quadrotor has been increasingly become popular. The application for this type of UAV has also become numerous, ranging from exploration, military missions, and search and rescue. Normally, a Proportional Derivative Controller (PID) controller is used for controlling the altitude of a quadrotor, because of its simplicity. However, the PID controller has a huge disadvantage when dealing with payload drop capabilities. A Gain Scheduling PID (GS-PID) controller can be used to reduce this overshoot, however, the current limits of a GS-PID controller is its restrictions to only states before payload drop and on the payload drop occurrence itself. In this study, a Fuzzy GS-PID controller is proposed for widening the GS-PID controller for considering varying payload weights. The study was able to design a quadrotor testbed for payload drop testing with a maximum payload of 300g. The testbed constitutes the quadrotor system, which was specifically designed for carrying heavy loads, and a height controller test rig, which is specifically designed to test the height controller of a quadrotor. In the testbed, numerous GS-PID controller were tested and analyzed on how the gains affected the overshoot of the quadrotor. One set of GS-PID gains are then used for comparison. After these tests, a Fuzzy GS-PID controller was designed and tested. Results show that the proposed controller was able to adapt to the varying payload weights. The results are then compared to one set of GS-PID controller and a PID controller in different payload weights.