Marine current turbines (MCTs) harvests kinetic energy from tidal currents in the ocean. In this work, Reynolds-Averaged NavierStokes (RANS) and unsteady Reynolds-Averaged NavierStokes (URANS) equations are used to study the flow physics at the turbine tip, and surrogate optimization models are used to improve the power coefficient of the MCT. The MCT was equipped with a winglet, and it was optimized for winglet height (h) and cant angle (alpha) to improve the power coefficient (CP). A large-scale optimization algorithm was used to find the optimal design in response surface method (RSM), kriging (KRG), radial basis neural network (RBNN), and weighted average surrogate (WAS), multiple surrogates enhance the performance and robustness of the optimization. The optimized MCT was compared with reference turbine experimental results and the optimized turbine improves the power coefficient by 10.46% for the cant angle of 33.18 degree and the height of 2.04% R. In this presentation, turbine design, optimization, and CFD simulation preliminary results and future work plans will be discussed.