Mechanical Engineering Faculty Publications

Comparative evaluation of adaptive and classical controllers for quadcopter trajectory tracking

Document Type

Article

Publication Date

2-7-2026

Abstract

Controlling quadcopters presents significant challenges due to their complex dynamics, strong nonlinear couplings, high system order, and underactuation. This paper compares the performance of several control strategies for trajectory tracking in quadcopters. Among these, four are optimal control approaches designed to approximate optimal solutions. The first two are based on approximate dynamic programming and include the adaptive critic and single network adaptive critic methods. The third is a model predictive controller, representing a rollout-based optimal control strategy. The fourth is a linear quadratic regulator. For baseline comparison, a conventional PID controller is also implemented. To benchmark performance, simulation results are compared against solutions obtained from a commercial optimal control solver, DIDO.

Comments

Reprints and permissions

https://rdcu.be/e7EMM

Publication Title

Soft Computing

DOI

10.1007/s00500-025-11046-4

Link to Full Text

Share

COinS