Hutsul T. Geoinformation multi-agent optimization of road planning (using Chernivtsi region as an example).

The planning of transport streams using meta-heuristics ant colony optimization method on based on the number of geoinformation agents is under exploration in this dissertation. Foreign experience in using ant colony optimization algorithms is analyzed in solving optimization tasks. Modern stare of investigation is also checked. Earlier ants were used for graphs and the results of such discoveries demonstrated optimal solutions. The existing requirements for planning the passing of the future route of the road network are determined. The sources of information for agents, the form of their representation for the mathematical model are substantiated. For practical verification of the hypothesis about the possibility of using the method of geoinformation multi-agent optimization in the planning of road networks, an appropriate software product has been developed. The road transport network of Chernivtsi Region is a topical research object. Differentiation of terrain conditions from plain to mountain, presence of dense river network, considerable forest cover, high level of economic development of the territory require special approaches to optimization and planning of development of the road transport network. However, the obtained results can be interpreted for any other territory of Ukraine. The directions of the future development of road transport networks have been clarified. The sources of information for the agents, the form of their representation for the mathematical model, are grounded. The practical test of the hypothesis regarding the possibility of using the method of geoinformation multiagent optimization in planning the development of road transport networks was carried out by the method of mathematical modeling. The parameters and peculiarities of the behavior of the indicated method under the various conditions and volumes of the input data are experimentally determined. The mechanism of processing the results of simulation and visualization of them in the traditional way is explained. On the example of the recommendations schemes the optimal ways of development of the road network of Chernivtsi region for the future are proposed and the possibilities of application of this method in the sphere of road planning are outlined. The result of the dissertation was a modification of the Max-Min method of multiagent optimization of the ant system for tracing the main and competitive variants of road transport networks, which is able to formulate various variants of passing optimal alternative paths between specified settlements.