No abstract

No abstract

No abstract

This study presents an artificial neural network (ANN) based adaptive proportional integral derivative (PID) controller algorithm which is developed to control the pitch angle of the vertical takeoff and landing (VTOL) system model. To find appropriate conventional PID controller parameters, either step response or vibration method might be used in a way of the widely used approach. They provide constant values for the conventional PID controller parameters for a closed-loop system however it is obvious that the unsatisfied tracking error performances in the closed-loop system might be addressed as a problem to be optimized. To overcome this problem, the parameters of the proposed adaptive PID controller might be determined with an ANN constructed feedforward multilayer perceptron. The proposed controller algorithm possesses the gradient descent with momentum update rule for the adaptiveness of the obtained PID parameters. The proposed adaptive PID controller algorithm is tested for the pitch angle of the VTOL system model in the MATLAB/Simulink environment in terms of the sinusoidal and step signals as the desired outputs. The obtained results are compared to the conventional PID controller whose parameters are tuned by Simulink PID tuner application in terms of mean square error, integral absolute error, the settling time, and the percentage overshoot.

No abstract