The estimation of railway track parameters from a target frequency response function is a non-convex optimization problem. The objective of this thesis is to evaluate new evolutionary optimization methods to solve the railway track optimization problem. We use a Python based platform, since this programming language is increasing in popularity and new evolutionary algorithms are being made available there. The purpose of the research is to find out which optimizer and objective function performs best. Before the railway track parameters are optimized, tests are conducted on benchmark problems to become familiar with the evolutionary optimization solutions. In this thesis we use Grey Wolf Optimization, Particle Swarm Optimization and Genetic Algorithm. Subsequently, the optimizers are applied on a numerical railway track model from [1]. We focus on the optimization of four parameters, stiffness and damping of both railpad and ballast. Objective function one J1 includes the sum of the differences between estimated frequency response function and target frequency response function in the frequency range 0-12500 Hz. Objective functions two includes the sum of the differences between estimated frequency response function and target frequency response function in the frequency range 0-3418 Hz and gives extra weight to those differences in the frequency range of the track resonances. Objective function three J3 includes the sum of differences of the estimated track resonances and the resonances from the target response. We compare the performance of these optimizers and three different objective functions. We show that it is possible to use all three optimizers to estimate the rail track parameters. We yield the best results with an objective function that only takes the frequency range 0-3418 Hz into account and applies a higher weight factor to differences within the frequency ranges of the track resonances. The compared algorithms behave differently for different objective functions. In most of the tested cases, GWO performed the best. The smallest difference between target parameters and optimized parameters were obtained with objective function two J2. Damping of the railpad is the most difficult parameter to estimate. Stiffness of the ballast was estimated with an error of about 3.90%, stiffness of the railpad with 1,06% and damping of ballast with an error of 3.60%. Finally, part of the further research includes the analysis of other evolutionary computation algorithm, optimization of the whole track model, sensitivity of analysis of the optimization parameters, inclusion of real-life measurements and addressing stochasticities in the objective function.