Optimized Neural Networks-PID Controller with Wind Rejection Strategy for a Quad-Rotor

Abstract

This study presents a comprehensive modeling and intelligent control strategy for quad-rotor aircraft, a four-rotor unmanned aerial vehicle (UAV). In fact, a PID on-line optimized Neural Networks Approach (PID-NN) has been created for controlling quad-rotor angular trajectories. PID classical controllers, on the other hand, are used to control position, altitude, and speed. The goal of this project is to design a smart Self-Tuning PID controller for attitude angle control that is based on neural networks and capable of supervising a quad-rotor for optimal behavior while tracking a desired trajectory. If the quad-rotor is navigating in hostile environments with irregular disturbances in the form of wind modeled and applied to the overall system, many challenges may arise. The quad-rotor must perform tasks quickly while maintaining stability and accuracy, and it must make decisions quickly in the face of disturbances. This method has a few advantages over traditional control methods like PID controllers. The results of the simulation are based on a comparison of PID and PID-NN controllers based on wind disturbances. These are used to test the quad-behavior rotor's and stability at various levels of strength. These simulation results are satisfactory and show that the proposed PD-NN approach is effective. In fact, the proposed controller has lower errors and a better ability to reject disturbances than the PD controller. It has also proven to be extremely resilient and effective in the face of turbulence in the form of wind disturbances.