Modelling and nonlinear control of a single rotor UAV

Abstract

Unmanned aerial vehicles have rapidly emerged as a promising technology for various industrial applications. UAVs can operate without a pilot, making them ideal for conducting tasks in dangerous or difficult-to-access locations. They can revolutionize agriculture, mining, construction, search, and environmental monitoring. Single Rotor UAV (SR-UAV) is an exciting alternative to conventional UAVs due to their lower weight and size. Moreover, they offer an efficient and practical solution for various applications requiring high mobility and maneuverability. Consequently, developing autonomous behavior depends critically on implementing suitable control systems. This is especially important considering that most UAVs are small and exposed to the effects of environmental phenomena such as wind, increasing the complexity of the problem. Therefore, effective control systems are essential to ensure the success of autonomous flight. This research work addresses the issue of modelling and nonlinear control of a single rotor UAV. The nonlinear dynamic equations are modeled for the translational and rotational motions. The control structure is based on a translation controller and a P-PI in cascade for the attitude controller, as well as the issues of robust controllers which are able to counteract external and unknown disturbances. Three control approaches have been selected and compared: classical PID, Super-Twisting (ST), and Adaptive Sliding Mode (ASM) for translational control. An automated tuning method using the particle swarm optimization (PSO) technique is presented to overcome the difficulty of tuning the parameters by trial and error procedures. An efficient tool, particle swarm optimization (PSO), is employed in this study to find optimal parameter values more quickly and accurately compared to conventional methods. The main objective is to enhance the performance and efficiency of the SR-UAV. The study aims to overcome the challenges associated with parameter tuning by utilizing the PSO technique. This approach enables fine-tuning the SR-UAV model with greater precision and effectiveness, ultimately leading to improved results. Using PSO as a tool for automated tuning facilitates optimization, reducing the time and effort required to achieve optimal performance. The controllers selected were simulated in scenarios with wind gust disturbances, and a comparison was made between the performance of the three controllers with and without optimized gains. The results show a significant improvement in the performance of PSO-tuned controllers. For all controllers, simulations were performed using tools such as MATLAB/Simulink. These simulations provided accurate data on the behavior of the controllers in different scenarios, allowing a fair and accurate comparison of their performance. The controllers were compared regarding accuracy using the tracking Root Mean Square Error (RMSE). For each of the simulation scenarios, the RMSE value was recorded for each controller.