Optimizing Highway Networks: A Genetic Algorithms and Swarm Intelligence Based Approach
Manoj K. Jha · 2002
Optimizing highway alignments between given end points is a complex problem since there are numerous alternatives to connect two points in space. The initial efforts to design a highway were primarily manual and based on human judgment. Later, a number of search algorithms were developed to find an optimum alignment after formulating a number of alignment significant costs such as earthwork and pavement costs. Genetic Algorithms (GAs) were recently used to optimize highway alignments and were proven to be very effective due to their ability to avoid getting trapped in local optima while searching for a global optimum solution. A Geographic Information System (GIS) was integrated with GAs for practical application, which allowed working directly with real maps. The genetic approach however, had two main weaknesses: (1) a predetermined number of generations through which the search for an optimal solution was to be carried out, was necessary. It was a usual practice to continue searching through sufficiently large number of generations to ensure that global optimal solution was reached, which increased the computational burden considerably, (2) the integration with GIS while allowed working directly with real maps, further increased the computational burden due to the additional computation necessary in the GIS environment. In order to address the computational burden issue, here we introduce an alternative approach using Swarm Intelligence (SI) for highway alignment optimization. We perform a test example which indicates that swarm intelligence reduces the computational burden significantly.