Generating future land-use and transportation plans for high-growth cities using a genetic algorithm

An elitist genetic algorithm was used to find a diverse non-dominated set of optimal future zoning and street plans for two high-growth cities in the United States of America. Plans were judged with regard to housing capacity, employment capacity, greenspace, traffic congestion, and change from the status quo. A multiobjective fitness function was used. The genetic algorithm offers the possibility of efficiently searching over tens of thousands of plans for a trade-off set of non-dominated plans. The trade-off set ranged from a minimum change plan, where undeveloped farmland was rezoned as commercial or residential land, to a minimum traffic congestion plan where commercial and residential usage were spread throughout the cities rather than concentrated in one or two areas. The algorithm is general enough to be applied to other cities and metropolitan regions.