Linear controller design with the use of PSO algorithm for UAV trajectory tracking

Abstract

This paper uses the PSO (Particle Swarm Optimization) algorithm to linearly control a BSP unmanned aircraft, enabling accurate tracking of a predefined trajectory. In the process of control optimization, the PSO algorithm was implemented with a quadratic cost function. During the study, unstable algorithm behavior was observed, as a result of which a modification was made by introducing a coefficient suppressing the movement of particles in the search space. The main task of the coefficient is to ensure the convergence of the solution. The effect of the coefficient value on population behavior was tested. In the simulation research on tracking, the previously generated trajectory was used, which is a reference for the location, velocity, and course for the linear dynamic BSP model. The results of these tests were presented in the form of waveforms of control signals and waveforms of state variables. By selecting the appropriate parameters, the proposed algorithm enabled the repeatability of results, as well as the proper mapping of the trajectory at the time of the operation of the algorithm not exceeding 0.03 seconds.

Comparison of optimization approaches on linear quadratic regulator design for trajectory tracking of a quadrotor

“…In [40] and [24], the weighting matrices of the LQR controller were determined by Artificial Neural Networks (ANN) and PSO; however, in both studies, the controllers were designed to control UAVs with three degrees of freedom. In [35], Multi-Objective High Exploration PSO was used to adjust the weighting matrices of the LQR controller for a nonlinear four degrees of freedom quadrotor model, while six degrees of freedom quadrotor models were used for the optimization of UAV trajectory tracking utilizing the PSO algorithm in [45,21].…”