In this thesis, we propose a methodology for choosing optimal mixed-level fractional factorial robust parameter designs when experiments involve both qualitative factors and quantitative factors. At the beginning, a brief review of fractional factorial designs and two-level robust parameter designs is given to help understanding our method. The minimum aberration criterion, one of the most commonly used criterion for design selection, is introduced. We modify this criterion and develop two generalized minimum aberration criteria for selecting optimal mixed-level fractional factorial robust parameter designs. Finally, we implement an effective computer program. A catalogue of 18-run optimal designs is constructed and some results are given. / February 2016