machine learning framework
examples

A generation of Cartesian robots with high speed requirements shall be optimized in its stability behavior over the working space.

For the optimization of velocity and acceleration models, the dynamic behavior was measured in various scenarios over the work space. Signals representing vibrations and other parameters were transformed.

From the resulting parameter space a fuzzyfied decision tree and fuzzy rules were created. To calculate concrete velocity and acceleration set up values, numerical optimization was applied to create computationally accurate and robust fuzzy rules.

The shown graph represents a rule base, which describes a vibration function over a special parameter space.