The objective of this project is to develop an algorithm able to obtain the optimal distribution of Dynamic Vibration Absorbers (DVAs) in underground railway infrastructures for reduction of railway-induced vibrations. The scope of the project is limited to the application of DVAs on the tunnel's walls. The Green's functions of the superstructure/tunnel/ground/building system and the dynamic response of the train will be input parameters of the algorithm and they will be provided by the research group.

This study develops an algorithm in order to obtain the optimal distribution of Dynamic Vibration Absorbers (DVAs) in underground railway infrastructures for the reduction of railway-induced vibrations. It has been developed and formulated the railway-induced vibration problem to obtain the vibrational response of the system. With this theoretical background a genetic algorithm (GA) has been implemented and refined; with the aim of saving time some processes have been applied to the GA, parallel processing, as well as vectorization of the code are the most successful methods. Furthermore, to release computing charge of the GA, a hybrid algorithm has been proposed, combining the GA with a pattern search algorithm to improve the accuracy of the variables of type double. Finally, to ensure that the solutions obtained correspond to global minimums, it is proposed a testing model using different software, OPTI Toolbox Studio to check the solutions.