Expected value model for optimizing the multiple bus headways

This paper studies the optimal multiple headways determination for a single bus route with a stochastic expected value model. The stochastic events from the model consist of random passenger arrivals, random boarding/alighting as well as random bus travel time. The expected value model employs two objective functions to maximize the bus company profit and to minimize the passengers waiting time cost respectively. The model is solved by a hybrid intelligent algorithm in which stochastic simulation and genetic algorithm are used to treat with the uncertain function. The algorithm efficiency is tested by a numerical experiment and concerned conclusions are drawn.

► Uncertainties of a bus system are taken into consideration.
► A stochastic expected value model is developed to calculate optimal bus headways.
► The proposed model is solved by combing stochastic simulation into genetic algorithm.
► The headways strategy which has lowest waiting time cost would be passenger’s expectation.