A mass transit system multicriteria frequency model solved by an interactive search procedure

A multicriteria frequency model for a mass transit system is formulated in the paper. The model contains several criteria which are important both for passenger and operator. The model tends to find optimal operating frequencies of vehicles in the mass transit network. It is formulated in terms of multiobjective, non-linear mathematical programming. The model is solved by an interactive procedure called Light Beam Search method, characterized by phases of computation alternating with phases of decision. At each computation phase a sample of efficient (non-dominated) solutions is generated. Those solutions are examined by the Decision Maker (DM) in the decision phase. As a result of the analysis the DM can input some preferential information which tends to improve the proposals generated in the next computation phase. In the analyzed case the decision consists in finding operating frequencies on all routes of the mass transit network. Those frequencies are the compromise solutions from the point of view of the following criteria: headway, waiting time, transferring frequency, degree of crowdedness, effectiveness and operating costs.