Optimal fleet size distribution and scheduling of transit systems using genetic algorithms

Optimal fleet size distribution and scheduling with transfer consideration for a transit system is a difficult optimization problem. A traditional formulation of the problem leads to a large non‐linear mixed integer programming problem. Past experience has shown that traditional optimization methods are unable to give optimal solutions to even simpler versions of the problem (like the optimal scheduling problem with known fleet size distribution). In this paper, a simple binary coded genetic algorithm (GA) based approach to the optimization problem is presented. The use of GA allows a more efficient formulation of the problem and the GA based approach gives optimal/near‐optimal results with limited computation effort.