Session: 17-01-01: Research Posters

Paper Number: 142473

142473 - Prediction of Friction Coefficient of Forklifts by Genetic Algorithm and Long Short-Term Memory 

Forklifts are used widely in various industrial fields. Forklifts on the road can be rollover in the pitch direction because of its load condition. A counterweight on the backside of forklifts compensates disproportionate load distribution occurred by the load on the fork, and it prevents rollover of forklifts. However, when the forklifts decelerate in an emergency, forklifts have a possibility to rollover. Moreover, when the load condition become more worse, they can tip-over about the tip of the fork. Therefore, it is important to predict the rollover of forklifts for safety.

This study aims to predict the longitudinal motion of the forklifts using vehicle dynamics theory and Pacejka’s tire model so that the possibility of rollover of forklifts can be predicted. Experiments needs to be progressed to measure the friction coefficient of tires which defines the tractive force of forklifts. Instead of additional experiments, the two methods: genetic algorithm and long short-term memory are adapted to predict the friction coefficient. A mathematical model with Pacejka’s tire model is built with the two methods, and the response of forklift’s longitudinal motion is compared to the simulation results for validation.

Keywords: Forklift, Genetic Algorithm, Long Short-Term Memory, Multi-body Dynamics Simulation, Tip-over

Presenting Author: Seungwoon Park Inha University

Presenting Author Biography: 2015.03-2022.02 B.S.Student (Inha University, South Korea)
2022.03-preset M.S.Student (Inha University, South Korea)
2022.03-2023.02 Visitor scholar (San Jose State University, United States of America)
Research interests: Multi-body dynamics, Vehicle dynamics, Structural analysis, Control

Authors:

Seungwoon Park Inha University
Chul-Hee Lee Inha University