Balancing Enhancement of the Ball on the Plate System Based on Pseudo-PD Controller and Machine Learning Algorithms.

EasyChair Preprint 5973, version 2

Abstract

This paper presents a novel method to improve the stabilization and trajectory 
tracking of the ball on the plate system (BOPS) based on machine learning algorithm 
with  the  Pseudo  proportional-integral-derivative  (PPID)  controller.  The  proposed 
controller depends on a machine learning (ML) algorithm that detect the angle of the 
servo motor required to correct the position of the ball on the plate. This paper presents 
three different ML algorithms for the servo motor angle prediction and achieved higher 
accuracy which are 99.85%, 100%, and 99.998% for support vector regression, decision 
tree regression, and random forest regression, respectively. The simulation results show 
that the proposed method has significantly improved the settling time and overshoot of 
the  system.  The  mathematical  formulation  can  be  obtained  using  the  Lagrangian 
formulation  and  the  servo  motor  parameter  obtained  by  a  practical  identification 
experiment.

Keyphrases: Ball on plate system, Pseudo-PD controller, machine learning, stabilization enhancement

@booklet{EasyChair:5973,
  author    = {Mohamed Elshamy and Mohamed Shoaib and Essam Nabil and Amged Sayed and Belal Abozalam},
  title     = {Balancing Enhancement of the Ball on the Plate System Based on Pseudo-PD Controller and Machine Learning Algorithms.},
  howpublished = {EasyChair Preprint 5973},
  year      = {EasyChair, 2021}}