Approach for Optimizing 3D Highway Alignments Based on Two-stage Dynamic Programming

Abstract

Optimizing highway alignments is a very complex engineering problem. None of existing methods has totally solved the problem In this paper we build an optimization model for highway 3D alignment based on two-stage dynamic programming, in which the comprehensive cost and design constraints are embedded. In the first stage: we formulate the optimization of alignment as a network problem by establishing a three-dimensional grid representation for the study area. Then dynamic programming is used to find the best corridor alignment. In order to obtain a group of scenarios, we propose a bidirectional searching strategy in dynamic programming. In the second stage: we put forward an improved direction acceleration method considering constraints to refine the corridor alignments for resulting the alignment deviate from the grid points and reaching the global optimum. Experimental results show that the global optimal alignment with good diversity was ultimately obtained.