The problem of improving the quality and shortening the development and design of asphalt concrete mixtures for surfaces of public roads, as well as forest ones is considered. It is noted that the existing methods are not sufficiently operational and correct in cases where the original or intermediate settlement data is carried by the property of the uncertainties of various types. The problem is especially urgent at the present time after the introduction of new GOSTs, which define a large range of mixtures in terms of their compositions. Under these conditions, using traditional methods, the task of predicting the properties of asphalt concrete mixtures becomes a laborious, time-consuming procedure and insufficiently accurate. To solve the problem, new, more advanced methods of designing the main parameters of asphalt concrete mixtures are needed, which use modern advances in information technology. An important parameter in the design of the mixture is to determine the degree of compaction of the road surface, which is directly determined by the content of air voids in the mixture, which was the purpose of this work. The aim of the work is to create an intelligent system for determining the content of air voids in the compacted asphalt concrete mixture of the road surface. To achieve the goal, the following tasks are solved: 1) implementation of the formulation of the task of designing the main parameter of asphalt concrete mixtures of road surfaces; 2) justification of the input and output variables of the problem; 3) preparation of initial data for training samples; 4) development of a neural network for determining the air voids of the asphalt concrete mixture; 5) software implementation of an intelligent system in the Matlab environment. The result of the research is the developed neural fuzzy network for the selection of asphalt concrete mixture with the determination of the content of air voids and its software implementation in the Matlab environment. Practical application of the results is provided for the selection of asphalt concrete mixtures for road surfaces.
Translated title of the contributionNEURO FUZZY NETWORK FOR SELECTION OF ASPHALT CONCRETE MIXTURES BY AIR VOID CONTENT
Original languageRussian
Pages (from-to)78-85
Number of pages8
JournalСистемы. Методы. Технологии
Issue number1 (53)
DOIs
Publication statusPublished - 2022

    Level of Research Output

  • VAK List

    GRNTI

  • 00.00.00 SOCIAL SCIENCES IN GENERAL

ID: 29729584