Publikace detail

Swarm Intelligence Based Multiple Model Approach for Friction Estimation at Wheel-Rail Interface

Autoři: Onat Altan | Voltr Petr

Rok: 2017

Druh publikace: ostatní - článek ve sborníku

Název zdroje: 5th International Symposium on Engineering, Artificial Intelligence and Applications : full paper proceedings

Název nakladatele: Girne American University

Místo vydání: Kyrenia

Strana od-do: 187-194

Tituly:

Jazyk	Název	Abstrakt	Klíčová slova
eng	Swarm Intelligence Based Multiple Model Approach for Friction Estimation at Wheel-Rail Interface	On board diagnosis and condition monitoring become a primary concern for railway vehicles so that necessary actions can be taken before a catastrophic consequence. Especially, the main aim is to diagnose and monitor such conditions by using vehicle mounted sensors. In this study, a particle swarm intelligence based multiple model approach is proposed to detect friction condition changes at wheel rail interface. In order to validate the approach, a mathematical model of a tram wheel test stand is given and this model is validated by the measurements taken from the various sensors attached to the wheel and roller of the test stand. Then, dry friction condition is detected by using angular velocity measurements. In this approach, multiple models imply mathematical models with different maximum friction coefficient values which are initially distributed uniformly in parameter space. Instead of fixed models, in this approach models are evolving based on the particle swarm intelligence. This methodology can be used as an on board condition monitoring tool for traction vehicles or as an auxiliary system for traction control system.	swarm intelligence; multiple models; friction estimation; tram; roller rig; wheel-rail contact; parameter estimation

Search