Designing granular layers for railway tracks using ray optimization algorithm

In this paper, the recently developed meta-heuristic algorithm, named Ray Optimization (RO), was applied to optimize the thickness of granular layers in railway tracks. RO is a multi-agent algorithm, each agent of which is modeled as a ray of light that moves in the search space in order to nd global or near-global optimum solutions. To utilize RO, considering structural and serviceability constraints, rst, hypothetical values of three thicknesses were assigned to the layers. Then, stresses under the sleeper and on the subgrade layer were compared with the allowable values. The total minimum thickness of the layers was also compared with possible minimum thickness (serviceability constraint) and then minimum thickness values were selected. Optimization results showed that the moment of inertia of the rail had no signicant impact on the minimum total thickness of the ballast and sub-ballast layers. On the other hand, train speed and axle load were the parameters that had a considerable eect on the minimum total thickness of granular layers in the railway track.