Due to the presence of transportation infrastructure in metropolitan areas, authorities often seek to modify the capacity of existing urban transportation links. Therefore, to meet travel demands, it is necessary to consider public transport (PT) modes and carry out systematic planning. In this research, assuming the existence of urban transportation infrastructure, we develop a mathematical model to determine the amount of link capacity increase for the dual-mode PT network. Such an increase in capacity can improve the network capacity and travel time reliabilities under both normal and peak traffic conditions. To investigate this issue, travel demand is assumed to follow a lognormal distribution under normal conditions. Then, the total travel time and flow through the links are calculated and applied to the model using the Monte Carlo method. In this study, we assume that the links of both transport modes forming a path are interdependent. Moreover, due to NP-hardness, we propose the PSO algorithm with dynamic parameters to solve the multi-objective bi-level model. Considering two public transport modes, we develop the algorithm for a part of the urban transportation network in Tehran, Iran.
