Kinematic Gauge Optimization of Switches Using Genetic Algorithms

A methodology for minimizing rail profile degradation in the switch panel of a railway turnout (switch & crossing, S&C) through kinematic track gauge widening is presented. A holistic approach is applied using a genetic algorithm for multi-objective optimization, where both routes and both moves (travel directions) are considered simultaneously. The dynamic vehicle–turnout interaction and rail profile wear as a function of gauge configuration are evaluated using a multi-body simulation tool. In the obtained set of Pareto optimal solutions, the trade-off between wear in the two routes is highlighted. A stochastic spread in wheel profile and wheel–rail friction coefficient is accounted for.