Machine learning approaches for predicting oil performance loss in wet brakes and clutches

Institut

Lehrstuhl für Maschinenelemente

Typ

Semesterarbeit / Masterarbeit /

Inhalt

theoretisch /  

Beschreibung

Initial situation:

Reliable friction behavior is crucial for the shifting performance and comfort of wet brakes and clutches in limited-slip differentials, active transfer cases, multi-speed transmission systems, and disconnect systems. Various experimental test series have revealed that composition and contamination through wear-related iron particles or water of modern lubricants significantly affect the friction and NVH (Noise, Vibration, Harshness) behavior as this may cause damage to the entire transmission unit. Therefore, a methodology was developed to quantify the performance loss of the lubricant by the use of defined parameters.

 

Objective:

Within the scope of this thesis, an existing dataset will be preprocessed to enable the application of data-driven modeling techniques. Subsequently, a predictive model will be developed using machine learning algorithms and evaluated in terms of its predictive accuracy and efficiency, in comparison with models reported in the literature.

Voraussetzungen

• Interest in the field of drive technology

• Experience with ML / AI or data modeling desirable

• Ability to work independently and well-structured

Möglicher Beginn

sofort

Kontakt

Johannes Wirkner, M.Sc.
Raum: MW 2506
Tel.: +49 89 289 15844
johannes.wirknertum.de

Ausschreibung