PENALAAN MANDIRI KOMPONEN PID DENGAN METODE JARINGAN SYARAF TIRUAN PADA SISTEM KENDALI QUADROTOR; SELF TUNING PID COMPONENTS WITH ARTIFICIAL NEURAL NETWORK METHOD IN QUADROTOR CONTROL SYSTEM

RAHANI; FAISAL FAJRI, Andi Dharmawan

2016 | Skripsi | FMIPA

Quadrotor is one type of unmanned aircraft that has the ability to fly off vertically with the ability to fly in a stable condition. It takes a good control system to acquire such capabilities. In this study, designed and created a system that is capable of stabilizing quadrotor currently hovering flight conditions by maintaining the angle of roll, pitch, and yaw using PID and neural network. PID method is one of the common control system applied to the quadrotor. This study implements neural network method to obtain the PID parameter values of the angle of roll, pitch, and yaw. Neural network method is used 2 input layer, 3 hidden layr, dan 1 output layer in the process of calculating value of Kp, Ki, and Kd. Results tuning PID using Ziegler-Nichols method and classical tuning has generated optimal value PID componens for each angle, is used as a starting point in determining the value of the component PID neural network the result of implementation of PID neural network control produce a stable control system with a value of steady-state error ? ± 2,4? roll angle, pitch angle of ? ± 2,5?, and yaw angle ? ± 3?.

Kata Kunci : UAV, VTOL, PID Ziegler-Nichols, 2-3-1 ANN