Enhancing trajectory tracking of Quadrotor using feedback linearization with MPC-LPV and LQI-LPV under variable disturbances

Abstract

The deployment of Unmanned Aerial Vehicles (UAVs) has markedly improved industrial efficiency. Optimizing control algorithms for precise path tracking is essential for enhancing the reliability and performance of these systems. This paper presents a comparative analysis of Model Predictive Controller (MPC) and Linear Quadratic Integral (LQI) controller, combined with the Linear Parameter Varying (LPV) approach, for quadrotor trajectory tracking under variable disturbances. The quadrotor is modeled using Newton-Euler equations for a six-degrees-of-freedom system. The control architecture includes an outer loop position controller using Feedback Linearization (FL) and an inner loop attitude controller implemented as either MPC-LPV or LQI-LPV. Disturbances are modeled using a custom sinusoidal model. Numerical simulations are done to evaluate controller performance for six different trajectories. The results for the helical trajectory indicate that under variable disturbances, the MPC-LPV+FL outperforms the LQI-LPV+FL. The MPC-LPV+FL achieves a steady-state error of 0.1934 m compared to 0.6611 m for the LQI-LPV+FL. Additionally, the MPC-LPV+FL reaches the minimum error in 3.1 seconds, compared to 3.8 seconds for the LQI-LPV+FL, resulting in a 22.58% faster response. Therefore, MPC-LPV+FL is the superior choice based on accuracy, stability, and speed of response under adverse conditions.