Optimization of the plasma discharge can be defined as determination of an optimal time evolution of the plasma parameters to lead a plasma to a desired state keeping it within the specific limits: physical ones (like the Greenwald density limit, low normalized beta and internal inductance values) and technical ones (like the vertical stability limit). The parameters, time-trajectories of which have to be optimized, are the ones significantly changing the plasma state, and, depending on the optimization goal, can be chosen from a wide range of plasma parameters: plasma current, plasma elongation, EC, NBI heating or current drive power, electron density, etc. Developing non-disruptive termination scenarios is important for safe operation of future tokamaks and especially for ITER since significant heat fluxes to the wall are expected during disruptions because of large amount of energy stored in burning plasmas. Therefore, the main goal of ramp-down optimization is to ramp down a plasma current as fast as possible while avoiding any disruptions. The results of the optimization problem study with the physical and technical limits is presented for TCV and AUG plasmas. The present work was done mainly with the RAPTOR code. The transport model has been extended to include a time-varying plasma equilibrium geometry, increasing the accuracy of full discharge simulations. Due to the design, the RAPTOR code is also an efficient tool for an optimization problem solving. A new ad-hoc transport model has been implemented to the RAPTOR code and tested during this work. Verification of the thermal transport model with simulation of the AUG and TCV full plasma discharges using RAPTOR will be presented.