The thinning of the silicon wafers and the smoothening of the surface are carried out by grinding and lapping processes. The lapping process is especially preferred to produce less surface damage of the silicon wafer in the production of high-efficiency solar cells. In this process, the surface roughness of the sample is the most influential parameter of pressure, rotation speed, lapping time and solution characteristics. In this study, rotation speed and lapping time were determined as design variables in the lapping processes of wire-sawn silicon wafers of 105 mm in diameter using a lapping and polishing machine. Although there are many studies about surface preparation in lapping process, the studies investigating this issue with theoretical or statistical modeling are very few in the literature. In addition, as a similar engineering process, many studies on the cutting process are available including experimental, modeling and optimization sections, which provides a good realistic design and prediction capability for the phenomenon. A similar study also needs to be done for the lapping process, which is especially preferred for the preparation of the surfaces of sensitive materials. The present study was conducted to fill this gap. In order to optimize the lapping parameters with a success, firstly, multiple nonlinear regression analyzes of experimental data were performed in terms of process parameters. Secondly, optimization studies were carried out based on Differential Evaluation, Nelder-Mead, Random Search and Simulated Annealing algorithms with the proposed regression models. It can be concluded that the present paper introduces significance of collaboration on the surface roughness experimental-modeling-optimization triple in silicon wafer lapping process for photovoltaic applications.