Bachelor/Master Thesis

Roller chain drives are widely adopted industrial power transmission systems, whose inherent polygonal effect generates periodic kinematic fluctuations, vibration and dynamic load variations. Traditional contact-based measurement methods rely on mounted sensors that alter the original chain operating state and limit full-field motion observation. Vision-based image acquisition enables non-intrusive capture of continuous chain movement, yet systematic pipelines to extract precise displacement, velocity and acceleration from chain test bench imagery remain insufficient. This thesis develops a complete image processing workflow to retrieve full kinematic quantities from roller chain test images, then builds a corresponding chain dynamic model and verifies its accuracy against the vision-extracted experimental kinematic data.

Bachelor/Master Thesis

Image-Based Kinematic Extraction and Dynamic Modelling of Roller Chain Drives

Tasks

  • Review existing image processing and optical flow motion measurement methods for mechanical component motion analysis
  • Preprocess and analyze raw image sequences captured from the roller chain experimental test bench
  • Develop a dedicated image processing pipeline to extract chain motion metrics: link displacement, instantaneous velocity and acceleration
  • Characterize the dynamic kinematic behaviour using the vision-derived motion data
  • Establish an analytical dynamic model of the roller chain drive and validate model outputs against the extracted experimental kinematic measurements
Start As of now
Software Matlab, Python
Requirements
  • Solid foundation in mechanics and data analytics
  • Ability to work independently
  • Enjoyment of programming tasks and strong motivation to learn
Beneficial TM I/II/III, Machine dynamics, GPU, Pytorch 

Further Information Dr. Junyu Qi (junyu.qi∂kit.edu)

 

Office Hours Monday/Friday 10:00 – 16:00, R. 706, Building 10.23

 

Literatures:

[1] J. Lee, M. Shinozuka: Real-Time Displacement Measurement of a Flexible Bridge Using Digital Image Processing Techniques. Exp Mech 46, 105–114 (2006).

[2] J. Guo, X. Wu, J. Liu, T. Wei, X. Yang, X. Yang, B. He, W. Zhang: Non-contact vibration sensor using deep learning and image processing, Measurement, Volume 183, 2021,109823

[3] S. Lin, S. Wang, T. Liu, X. Liu and C. Liu: Accurate Measurement of Bridge Vibration Displacement via Deep Convolutional Neural Network. IEEE Transactions on Instrumentation and Measurement, vol. 72, pp. 1-16, 2023, Art no. 5020016, doi: 10.1109/TIM.2023.3291786.

[4] Xu F. Accurate measurement of structural vibration based on digital image processing technology. Concurrency Computat Pract Exper. 2019;31:e4767. https://doi.org/10.1002/cpe.4767

[5] K. Son, H. Jeon, J. Park, J. Park: Vibration displacement measurement technology for cylindrical structures using camera images, Nuclear Engineering and Technology, Volume 47, Issue 4, 2015, Pages 488-499