On replacing a PID controller with Neural Network controller for Segway

Abstract

Segway is personal short-haul transportation. currently, the Segway moving that is available in the market can be moved by the user by moving his balance backward and forward. Due to the services provided by this machine, it is important to study its performance and find appropriate techniques to control it. The most well-known and famous controller in industrial applications is the conventional PID controller. It uses three functions (proportional, integrated, and derivative) to control the input signals before entering the plant or process unit. In this paper, another control method based on Neural Network (NN) control is tested to replace the PID controller to control the cart position and handlebar angle of the Segway. Simulation in this work is performed using MATLAB, these results indicated that the neural network controller used to control and identify the Segway gave a better performance and results compared to the traditional PID controller.