Multi-objective path finding in stochastic time-dependent road networks using non-dominated sorting genetic algorithm

• We propose a multi-objective path finding model.
• We consider the stochastic and time-varying nature of travel time in our model.
• The model is solved by the non-dominated sorting genetic algorithm.
• The Taguchi method is used to tune the parameters of the genetic algorithm.

In this paper, a new multi-objective path-finding model is proposed to find optimal paths in road networks with time-dependent stochastic travel times. This study is motivated by the fact that different travelers usually have different route-choice preferences, often involving multiple conflicting criteria such as expected path travel time, variance of path travel time and so forth. However, most of the existing studies have only considered the expected value of path travel time as the sole decision criterion. In order to solve the multi-objective model, the non-dominated sorting genetic algorithm is employed and its parameters are tuned by the Taguchi method. Moreover, a dynamic n-point crossover operator is developed to enhance the search capability of the genetic algorithm. Experimental results on a grid network demonstrate that the proposed approach is able to provide a set of non-dominated paths from which travelers can choose their paths based on their attitudes toward travel time uncertainty. Statistical analysis confirms that the dynamic n-point crossover operator outperforms the traditional one-point crossover operator.