OPTIMIZATIONAL TASK SOLUTION OF STATIONARY POINTS PLACEMENT FOR OBSERVATION OF ATMOSPHERIC POLLUTION AT TECHNOGENICALLY LOADED TERRITORIES OF UKRAINE

Abstract

Network of air pollution monitoring stations in Ukraine was built in the 1970s in accordance with the standards of the former USSR. Their con-figuration was not revised. During this period there were many changes in economy, industry, transport infrastructure, climatic conditions. It led to radical redistribution of technogenic load on air of Ukraine. Therefore, the existing network of posts today is no longer optimal. It does not allow to see real picture of pollution. This, in turn, does not make possible to make effective decisions on air quality management and public health risk in urban areas. This situation does not meet the pan-European requirements that Ukraine should comply with the Partnership and Cooperation Agreement between the European Union, the Member States and Ukraine. The adopted normative legal acts of Ukraine that one of the prior tasks of the existing air monitoring system development is theoretical substantiation and proposals preparation of optimized schemes for construction and operation of observation networks according to European requirements and standards. Therefore, development of mathe-matical tools for optimization problem solution of stationary points placement for observation of atmospheric pollution at technogenically loaded territories is an urgent scientific problem. Comparative analysis of different approaches to determining spatial configuration of the air monitor-ing network was identified. Their main shortcomings are identified. It makes almost impossible to use them in today's Ukraine. Mathematical formalization of optimization problem solution of stationary points placement for observation of atmospheric pollution at technogenically loaded territories is carried out. From the point of view of optimization theory, the obtained problem is dynamic, nonlinear, deterministic and discrete on a nonconvex domain. Due to considerable complexity of the problem, its solution (finding the optimal solution) is possible only by the method of complete search. However, application of this method is complicated due to the very large number of computational operations for large zones and agglomerations. So, there is a need to use new optimization algorithms. Two algorithms for optimization problem solving were developed. They are based on combination of greedy algorithm and complete search method. Testing of these algorithms (on the example of data from Kyiv) showed that they allow to obtain problem solution (close to optimal) much faster than the method of complete search.