To maintain the efficiency of road transport, urban road networks are supposed to meet the ever changing transport needs of the population. The most efficient way to reach rational solution in this area is using the evolutionary approach that treats a city as a self-organizing complex system. The study proposes a novel method of evolutionary modification of road network structure. This method is based on a combination of clustering and memetic algorithm. The clustering stage generates subgraphs of road network according to different spatial and semantic parameters that allow considering a lot of features of real objects in a compressed form. A memetic algorithm finds optimal road network structure using obtained clusters. The study combines two approaches to develop a road network (the urban planning and the transport engineering). Therefore the fitness function of a memetic algorithm contains the following criteria from the both approaches: a connectivity, a pedestrian accessibility, a provision of the territory with transport system, a reliability, and an accident rate. The method was verified on two datasets: the city of Samara and Koshelev Project. In both cases simulation has shown significant improvements of the optimized road network in comparison to existing one.
Transport infrastructure optimization method based on a memetic algorithm
2017-10-01
471824 byte
Conference paper
Electronic Resource
English
SAR image processing by a memetic algorithm
IEEE | 2005
|Automotive engineering | 2015
|A memetic algorithm for the max–cut problem
British Library Online Contents | 2015
|A memetic algorithm for the max-cut problem
British Library Online Contents | 2015
|