TOMSK, Jan 19 – RIA Tomsk. Scientists of the Institute of Applied Mathematics and Computer Science of Tomsk State University (IAMCS TSU) have created a mathematical model that can prevent equipment breakdowns at industrial enterprises and will be useful for oil and gas, petrochemical, power generating industries, the university's press service reported on Tuesday.
According to the press service, employees of the TSU Institute of Applied Mathematics and Computer Science with the support of the Russian Foundation for Basic Research, have developed a prototype of a process data analysis system - a mathematical model that can recognize normal and abnormal operation of equipment at industrial enterprises. The information for the formation of data was provided by the industrial partners of the university.
"The prototype finds any atypical fragments. For example, a mathematical model captures sharp bursts of signal amplitude, this is typical for electrical indicators - current, voltage; frequency change, for example, the occurrence of additional vibration in the presence of mechanical defects on bearings", - one of the authors of the project Damir Murzagulov is quoted.
If the system finds any abnormality, a notification will appear in the graphical interface. Then the operator will decide whether there is a threat. If it is not there, the specialist will mark this situation in the program as standard, and the next time it appears, the system will identify it this way. That is, during operation, the model is also trained.Timely recording of anomalies in the operation of the equipment will help prevent its breakdown and avoid downtime and economic losses. In addition, the mathematical models embedded in the prototype are not intended for a specific type of equipment, signal or production, and the model extracts all knowledge from historical data. Therefore, the prototype can be used in any production.
The TSU press service added that such a tool is needed by enterprises that use complex equipment, the emergency repair of which is very expensive. These are oil and gas, petrochemical, power generating enterprises. Now the world's leading companies are working on the design of such systems.