Add to favorite
 
123
Subscribe to our Newsletters Subscribe to our Newsletters Get Daily Updates RSS


Efficient Method of Teaching Electronic Noses Developed in Russia
May 29, 2018 15:34


Scientists of the Gagarin Saratov State Technical University together with colleagues from the Karlsruhe Technology Institute (Germany) and the Manas University (Kyrgyzstan) have developed an efficient way of training electronic noses. The results of the study are published in The Sensors.
Electronic noses are compact devices for the recognizing the smells of a wide range of chemicals. Unlike simple gas sensors, they are capable of not only distinguishing mixtures, but also memorize new smells through training with artificial intelligence systems. For 30 years scientists have been trying to create efficient electronic noses. The basic idea is not to determine which odor molecules appear in the air, but just to take a set of gas sensors that react to the appearance of certain impurities in the air. And afterwards "artificial intelligence" is used - the computer learns to recognize particular smells.
Training of the electronic nose is similar to the training of search dogs: a computer equipped with sensors is demonstrated certain smell to and imprinted under a specific label. For example, the sensor is aimed at searching for the smell of a rose in the atmosphere. The computer memorizes what kind of "image" a rose has and subsequently distinguishes it, for example, from the smell of sauerkraut.

But, unlike the biological smell system, the electronic nose, when placed in real conditions, with a mix of uncontrolled impurities arising in the surrounding atmosphere, begins to "get confused" in a few days. Scientists, often dealing with gas sensors, of course, know what it is about. "The fact is that the sensors are unstable and their readings drift with time," one of the project authors, Victor Sysoev explained.  "Therefore we need stable sensors that avoid drift. We have demonstrated how to solve this challenge with an electronic nose based on a multisensor metal oxide matrix of.”
Stable high-sensitivity sensors based on single-crystal nanofibers have been developed at the Gagarin Saratov State Technical University. But it turned out that electronic noses on their basis can accurately detect smells for only a week or two, and then require recalibration. Therefore, the University specialists tried to take a fresh look at this problem. It is all about a fast algorithm for recalibrating sensors with the possibility of fine tuning in the process of device’ operation.

"We have found out that the stability of the electronic nose depends mainly on random changes in the air composition, and the introduction of the nose training procedure ensures its stability for many years. Moreover, we have proved the ability of the device to function for more than eight years, - no one has ever demonstrated such a long and successful operation of artificial noses before,” - Victor Sysoev added.
This development is of great importance for practitioners, since it can be used to create an "ideal" electronic nose, which is a stable and simple technical tool for analyzing the composition of gas media.




Author: Vera Ivanova

Tags:      

Next Previous

You might also find interesting:

The Victory Tastes Sweet. Then What? Tomsk Scientists Warn about Threat of Global Climate Disaster Enzymes Are Able To Purify Water Blood Glucose Level Can Be Measured Through Blood Pressure New Cancer Medicine Found in Siberian Cave









Comment on our site


RSS   twitter   facebook   submit

Bookmark and Share

search on the map
TAGS:
Rubtsovsk  Cyber Fraud  New Films  Dima Yakovlev Law  Harold Feinstein  Mercedes-Benz Fashion Week Russia  TykvaDAY  Russian business  Public transport   Russian regions  Far East  Festivals in Moscow  Russian tourism  Russian engineers  Bone Carving  Exhibitions in Moscow  Russian airlines  Moscow  Alexander Lebedev  Chelyabinsk  TNK-BP  Republic of Tatarstan  travel to Russia  Zoo  prices  Russian musicians  foreign passport  Odyssey Festival  Alexei Kudrin  Armen Dzhigarkhanyan  Comics  Old Russian Icon Painting  Elections 2012  Yandex  Kaluga Region  buy Russian designers  City Hunter  Prison  Russian Cinema  Darwin Museum  Russian media  St. Petersburg  Burial Grounds  Chess Museum  Hans Feurer  gambling in Russia  Belokurikha  Vladimir  Russian academy of sciences  Big City Life 


Travel Blogs
Top Traveling Sites