Session: 07-02-02: Nonlinear Dynamics, Control, and Stochastic Mechanics

Paper Number: 140388

140388 - Many-Objective Optimization of Hybrid Control Systems for Quadrotor Unmanned Aerial Vehicles 

This paper introduces an innovative cascade control structure for quadcopter Unmanned Aerial Vehicles (UAVs), leveraging a combination of linear and nonlinear control strategies to enhance position and attitude management. At its core, the study develops a mathematical model of the UAV, employing both a nonlinear approach for the application of sliding mode control and a linear approximation for the design of specialized controllers. These controllers, identified as the X^G and Y^G controllers, are crucial for calculating the precise roll and pitch movements required to navigate the UAV to specified global positions, as well as for determining the rotors' desired angular rates, assuming minimal Euler angles.

 

The research progresses to tackle a multi-objective optimization challenge, defined by 21 design variables and 10 objectives, aimed at optimizing factors such as the UAV’s response speed, overshot, tracking error, and energy consumption. To solve this complex problem, the HypE (Hypervolume Estimation) algorithm, a notable many-objective optimization technique, is employed under strict performance and stability constraints. The optimization results are presented through Pareto sets and fronts, illustrating the competitive nature of the design objectives, and highlighting the intricacies involved in achieving a balanced solution.

 

Numerical simulations based on the nonlinear model of the UAV validate the proposed control structure's effectiveness, showcasing its superiority in managing the UAV’s dynamics. Furthermore, the paper provides a comprehensive overview of the current state of control engineering for UAVs. This includes a detailed examination of existing literature, comparing the efficacy of different control strategies, and emphasizing the importance of customization in control system design to accommodate the unique challenges presented by UAV dynamics.

 

The discussion extends to intelligent control mechanisms, such as fuzzy adaptive PD controllers and Neuro-Fuzzy controllers, highlighting the role of computational intelligence in advancing UAV control systems. This study not only demonstrates the feasibility of integrating linear and nonlinear control techniques in a cascade structure for UAV optimization but also underscores the value of employing a many-objective optimization framework to navigate through conflicting design objectives.

 

In conclusion, this paper makes a significant contribution to UAV control system design, offering a novel approach that combines detailed mathematical modeling with advanced optimization techniques to achieve optimal control performance. By addressing the complexities of multi-objective optimization and validating the proposed model through simulations, the research sets a foundation for future exploration and development in UAV control systems, emphasizing the necessity for continuous innovation in the face of evolving technological and operational requirements.

Presenting Author: Yousef Sardahi Marshall University

Presenting Author Biography: Dr. Yousef Sardahi is an Associate Professor in the Mechanical and Industrial Engineering Department at Marshall University. Dr. Sardahi earned a Ph.D. from the Department of Mechanical Engineering at the University of California, Merced, in 2016. His research interests include control system design and multi-objective optimization. He brings his wealth of knowledge and experience to the classroom, teaching a diverse range of subjects, including control systems, digital controls, automation and control, system modeling, advanced vibrations, mechatronics, circuits and instruments, and mechanical engineering computation. With his passion for educating the next generation of engineers, Dr. Sardahi is a true asset to the academic community at Marshall University.

Authors:

Xinhuang Wu Marshall University
Yousef Sardahi Marshall University