Design of Advanced Control Algorithm for Quadcopter

Lee, Hao-Ern (2020) Design of Advanced Control Algorithm for Quadcopter. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text LeeHao-Ern _Full Text.pdf Restricted to Registered users only Download (3MB)

Abstract

This research was to come out with a better performance control algorithm for the X-type quadcopter. Proportional- Derivative (PD) is the most common used control algorithm for the quadcopter due to its simplicity and decent robustness. However, since the quadcopter has a non-linear dynamic, the PD controller can limit the performance of the quadcopter. For instance, the disturbance rejection of PD was just simply not there. There were tons of control algorithms around and more to be formulated, and there were some algorithms that were commonly applied on the quadcopter. In the literature review, these control algorithms were compared upon each other. After comparing all these algorithms, the PD controller was used as the benchmark to compare with the Second Order Sliding Mode Controller (SOSMC) in terms of robustness, maximum overshoot, and settling time. A quadcopter simulation model was designed in this thesis. To increase the reliability of the simulation, the following subsystems was designed. Firstly, the input signal subsystem which served the purpose of providing an input signal to the flight control system. Next, is the sensor subsystem, which consist of the Inertia Measurement Unit (IMU) sensor to obtain the current position and orientation of the quadcopter. Followed by the environment subsystem, which provide the environment information to the airframe and sensor. These information included temperature, air density, gravitational acceleration, and pressure. Then, the flight controller subsystem, where the control systems were implemented at. The input signal and sensor data was directed into this subsystem. The control system will process these data and signal to provide the motor command. Furthermore, the airframe, where the dynamic of the quadcopter was implemented. The motor command was directed into this subsystem to prompt the motor run. Last but not least was the visualization subsystem, which a Graphical User Interface (GUI) was design for the user to have a better visualization of the flying quadcopter. After designing the simulation model, the SOSMC and PD control systems were designed to produce output to the motor. Parameters analysis was carried out to fine tune the parameters of the control systems. The simulation was focusing on the hovering mode, which is the altitude and yaw control only. In order to compare the control systems more accurately, there were different type of input signal directed to the control system to test out their maximum performance. These input were slow-rate input and fast-rate input. The simulation results was then be tabulated and compared between the performances of the control systems. After series of simulation and comparison, the SOSMC was proven to have a better performance than the PD controller. During the altitude tracking, the maximum overshoot had decreased by 97.27% in the slow-rate input and 94.72% in the fast-rate input. Where the delayed settling time decreased by 70% in the slow-rate input and 77.5% in the fast-rate input.