adVances In transportatIon studIes

The available Highway Alignment Optimization (HAO) algorithms use either single-objective or multiobjective approaches. These algorithms consider different highway alignment costs along with restricted land use information such as forests, wetland, etc. as the primary objectives which are minimized in the process of highway alignment optimization. In single-objective HAO approaches, different highway alignment sensitive costs are formulated and added to obtain the total highway alignment cost in monetary value. The total highway alignment cost is used as the primary objective function in the single-objective HAO. This method, upon optimal search, yields a highway alignment with minimum total highway alignment cost. As part of the total highway alignment cost, the restricted land use is expected to be minimized in the optimization process. Whereas in the multi-objective optimization approach the highway alignment cost and restricted land use information are maintained separate and optimized simultaneously. This way, the multi-objective HAO helps to yield a set of Pareto-optimal solution with trade-off between highway alignment cost and the restricted land use information. The restricted land use information used in both optimization processes is derived from a geographical information system (GIS) database. It is cumbersome to express the highway alignment and associated objectives in mathematical formulations as required in classical optimization techniques. Hence, both methods use unconventional genetic algorithm (GA) based optimization approach. This paper compares single-objective HAO with multi-objective HAO and discusses their merits and demerits in a real-world example application.