Application of artificial neural networks for solving the problem of two-dimensional geofields reconstruction

Pavel A. Kakovkin, Aleksey A. Druki

Результат исследований: Материалы для журналаСтатьярецензирование

1 Цитирования (Scopus)


The relevance of the research is caused by the necessity to develop algorithms and software to increase the efficiency of the spatial analysis and two-dimensional geofields recovery. The main aim of the study is to develop the algorithms and software to increase the efficiency of two-dimensional geofields recovery on irregular grid point data; to carry out the experimental studies to determine the effectiveness of the developed algorithms and to compare them with the existing analogues. The methods used in the study. To solve the task the authors have applied the methods of artificial intelligence, methods of implementation of artificial neural networks and genetic algorithms; the committee methods are applied to solve the classification problems, mathematical modeling, probability theory and mathematical statistics with the help of software Visual Studio and MatLab. The results. The artificial intelligence methods were used for restoring geofields on irregular grid point data, as this area of research is one of the most intensively developing now. The algorithm based on artificial neural networks was developed to solve the problem. The algorithm is a sequence of actions. It consists of seven steps. The multiple neural networks of direct distribution, such as perceptron, which operate according to the bagging method, are used to restore geofield. The software application that allows solving the assigned task is developed on the basis of the proposed algorithm. The authors carried out the experimental study of the algorithm effectiveness and compared the results obtained with the results of the inverse distance-weighted method. The carried out studies shown that the results of the proposed algorithm operation are higher than the operating efficiency of the inverse-distance weighted method.

Язык оригиналаАнглийский
Страницы (с-по)35-43
Число страниц9
ЖурналBulletin of the Tomsk Polytechnic University, Geo Assets Engineering
Номер выпуска2
СостояниеОпубликовано - 2016

ASJC Scopus subject areas

  • Materials Science (miscellaneous)
  • Fuel Technology
  • Geotechnical Engineering and Engineering Geology
  • Waste Management and Disposal
  • Economic Geology
  • Management, Monitoring, Policy and Law

Fingerprint Подробные сведения о темах исследования «Application of artificial neural networks for solving the problem of two-dimensional geofields reconstruction». Вместе они формируют уникальный семантический отпечаток (fingerprint).