Balancing Control of Dual-Rotor Unmanned Aerial Vehicle Using Linear Controller

Tan, Jin Hong (2023) Balancing Control of Dual-Rotor Unmanned Aerial Vehicle Using Linear Controller. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text Full Text - TAN JIN HONG.pdf Restricted to Registered users only Download (52MB)

Abstract

The use of drones or unmanned aerial vehicles (UAVs) has become increasingly popular in various industries due to their ability to perform tasks that surpass human capability. This research focuses on bicopter, a type of drone with two rotors and propellers, and their balancing control. Balancing a bicopter is challenging due to external disturbances, particularly static and dynamic load variations, which can cause imbalance and lead to crashes. To address this, the objective of this project is to design an enhanced linear controller for balancing control of a dual rotor UAV and evaluate its effectiveness against existing control algorithms. Two types of controllers are considered: classical controllers (PID, PI, PD, and cascade) and advanced controllers (NPID and sliding mode). While the PID controller stands out among the classical controllers due to its simplicity, it lacks performance in the presence of load variation. On the other hand, the NPID controller is shown to be a preferable approach compared to sliding mode due to its simplicity in design and application, and its ability to overcome nonlinear problems. Both PID and NPID controllers are designed and applied to a self-developed real-time bicopter through the programming of an STM32 microcontroller. The stability of the designed PID controller is validated using bode diagrams and Nyquist plots, and a nonlinear function is added to form the NPID controller. Angular position signals from MPU6050 sensor are collected and analysed, and a holding jig is used to create five testing conditions, including no load, single static 50g load, multiple static 50g loads, and dynamic 50g load. The experiments were conducted in a closed environment. Settling time and RMS error are chosen as the performance indicators in this project. Based on the collected result, the settling time of the NPID controller improved by 58.62%, 13.53%, 10.71% and 4.57% respectively according to the different load conditions. On the other hand, in term of RMS error, NPID controller also outperforms PID controller in most of the loading conditions except static 50g load condition. The overall improvement or deterioration according to the loading conditions will be 42.12%, -6.65%, 1.83% and 35.27% respectively. This result shows that the NPID controller is performing better compared to the benchmark PID controller in term of settling time and RMS error. In overall, this research contributes to the development of better balancing control for bicopter, which can lead to improved performance and stability in real-world applications, specifically in application of drone in transporting loads. Last but not least, future works includes responsiveness improvement, increasing more disturbance scenario, identification of maximum allowable carrying load on the bicopter and NPID controller enhancement can be made to encourage and motivate the continuation of research on this topic.