Pedestrian Activity-based Network Design based on Multi-Objective Programming

Abstract

Recently public space designs for pedestrian gather attention. Pedestrian activities are complicated for modeling because pedestrians probabilistically behave on high-resolution networks, and we cannot ignore the aspect of time-use. In this study, we present an optimization problem of the policy of street space reallocation, considering its effect on pedestrian activities. The problem is assumed as a pedestrian activity-based network design problem. By determining the sidewalk width of each link, we maximize the total activity time or the expected maximum utility of the district. On the other hand, the capital cost should be minimized. For describing this trade-off relationship, we formulate our problem as a multi-objective programming and solve it by the network update method. As the results, we obtain clear frontiers and the inspiration for the connection strategy between urban facilities.