Abstract:We propose a highway network planning procedure by considering the topological and realistic constraints. On the demand side, attachment probability for each node pair is calculated by a radiation model. The network planning is formulated as a mixed integer nonlinear programming, to maximize the total attachment probability of connected node pairs. Edge merging is applied to make realistic highway connection. The proposed method bypasses the requirement of OD matrices, and multiple paths for any OD pairs can co-exist in the planned network, which is different from the traditional planning procedures based on four-step model and node importance method. A case study for a real-world region with 100 cities in Shaanxi, China is presented to test the validity of the proposed procedure. It is shown that proper highway networks can be generated under scenarios with different attachment probability thresholds and budget constraints.