Master / Bachelor thesis: Swarm algorithm for the adjustment of a textile machine
In T-EXDIZ the experimentable digital twin of a tufting machine was developed. The machine consists of several levers and rods that can be adjusted to change the movement of a needle, knife, and gripper.
Specifications define desired stroke and the desired interaction between the tools. However, changes of one rod or lever can have influence on multiple specifications. The difficulty is to find correct settings, which is normally done with expert knowledge and through various iterations. In T-EXDIZ the optimization was achieved through discretizing the solution space and then simulating and evaluating all possibilities. However, this becomes computationally expensive as the number of parameters or desired accuracy increases. For 4 parameters and 10 discretized values each, we already must carry out 10^4 = 10,000 simulations.
As an alternative, in this thesis the student should implement a swarm algorithm to adjust the parameters of an existing digital twin of the tufting machine and compare the results with the existing algorithm.
Programming knowledge required.
Keywords: swarm algorithms, optimization, digital twin
Betreuer: Hüsener
