Optimization of the plasma discharge is related to determination of the optimal time evolution of the plasma parameters to reach a specific plasma state taken into account certain physical and technical constraints. The set of plasma parameters which should be optimized consists of the parameters significantly changing the plasma state and can be defined from the different tokamak actuator inputs: plasma current, EC, NBI heating or current drive power, density, etc. In this work we carry out the optimization study of the current ramp down. Numerical optimization of this phase can prescribe evolution of the plasma parameters to terminate plasma as fast as possible and in the same time to avoid disruptions in the real experiments. The simulation is performed with the RAPTOR code. It is a light and fast transport code with a simplified transport model which includes transport equations for electron temperature and poloidal flux. However, this code was constructed assuming a fixed plasma equilibrium, whereas plasma geometry might change during ramp-down phase. Therefore RAPTOR has been extended to include time varying terms. In this way, time varying plasma geometry can be used in the optimization procedure and for example plasma elongation can be an additional parameter for the trajectory optimization. The results of the simulation with the extended transport model and optimization procedure of ramp-down phase of AUG-like plasma parameters are presented.