A PSO highway alignment optimization algorithm considering environmental constraints

In recent years, there has been a growing need to develop and delineate algorithms that quickly provide optimized highway alignments. The matter is very complex since it is dominated by several factors and engineers have to achieve different and relevant targets (such as minimization of construction and earthwork costs, compliance with environmental and right-of-way constraints, maximization of safety and comfort for users, conformity with geometric standards). The possible solutions of this problem are clearly infinite and only modern artificial intelligence techniques can really simplify and speed up the highway design process. In this study a search algorithm, based on a Swarm Artificial Intelligence technique (Particle Swarm Optimization method), to optimize highway 3-dimensional alignments, considering also environmental constraints, is proposed. This algorithm pertains to the minimization of a particular cost function, made up of different kinds of construction costs and some penalties, related to geometric and environmental constraints (geomorphologic, hydraulic and seismic constraints) and useful to avoid and discard incorrect alignments. Some specific operators, derived from Genetic Algorithms (GAs), were also introduced in the model for improving its efficiency and correcting inappropriate solutions. To test the model efficiency, through the MatLab © software, an original script has been developed. The topography of the study area was reproduced through a particular Digital Terrain Model (DTM) representation.