Перейти к основному содержанию
AkademIndex

Продукты

Для разработчиков

AkademBaseскороОткрытый API экосистемы
Латиница
Русский
Статья

APPLICATION OF EXPERT SYSTEMS FOR MEASURING THE HUMIDITY OF BULK MATERIALS

E. UljayevTashkent State Technical University, UzbekistanUtkirjon Murodillaevich UbaydullaevTashkent State Technical University, UzbekistanSh. N. NarzullayevTashkent State Technical University, UzbekistanO NorboyevKarshi Engineering and Economic Institute, Uzbekistan,M KulakovN YusupbekovB MuhamedovG'ulomov ShE UljaevU UbaydullaevTashkent State Technical University, UzbekistanNarzullaev ShS ErkinovV SemeykinE UljayevTashkent State Technical University, UzbekistanT ShU UbaydullayevT UmarovU MardonovO KhasanovShB OzodovaYusupovE NarzullaevF XudoyberdievHaydarovE UmarovU ShoazimovaKhE UlzhaevO NarzullaevNorboev
ABI

Аннотация

Measuring instruments for monitoring temperature, pressure, and other parameters are designed with large permissible errors in the range from 0.1 to 5 10%. In particular, for measuring instruments, express moisture measurements of bulk error tolerances are in the range of 0.5%. The reason for such tolerances is associated with the difficulties of the methodology for ensuring the development and development of measuring and accounting mechanisms, both the change in the internal parameters of the device and the external parameters of the environment. The validity of the relevance of the work performed is associated with the need to develop modern measuring devices, that is, with the use of expert systems (ES) and artificial neural networks (ANN), which make it possible to create intelligent measuring devices. The work carried out the implementation of ANN in the mathematical package MATLAB 6.1, which shows the method of using an artificial neural network to create an intelligent device for an express method for measuring the moisture content of bulk materials, using the example of grain. After completing the preparatory stages, a neural network for measuring the moisture content of bulk materials was developed. As a result of the experiments carried out with three input parameters and updating the initial conditions, a learning curve was built that satisfies the main goal of learning ANN.

Темы

Идентификаторы

Цитирования и источники

Показатели — AkademScholar · Скоро