Friday, September 17, 2021, 11:00
The aim of this thesis is to focus on building a model of a production line and optimize it. In such a production line electronics controls are assembled by placing and soldering components on PCBs, short for printed circuit boards. Before the „surface mounted device“ components are placed on the PCB, the solder paste has to be printed on the board. This step is one of the most important steps in this production line. This step has many influencing factors, like temperature, humidity, printing speed or the cleaning cycles of the stencil. There are different sensors on the production line which gives us information about this data. Also testing data is available. On the one hand there are pass/fail results of optical tests and on the other hand there are electrical measurements with continuous numbers. It is tried to set up a classification model and also a regression model is set up. Later these models are optimized to improve input parameters and reduce errors. Apart from linear regression modelling, different machine learning approaches are tried, too.
Meeting-ID: 627 9173 7415