Intelligent diagnostic system for metallurgy stand mill
Intelligent diagnostic system for metallurgy stand mill based on the Digital Twin concept using Speedgoat Performance real time target machine
This project based on the Digital Twin concept, the purpose of which is an on-line diagnostic system of the electrical and mechanical parts of the stand mill 1700. This project is in line with the European direction for the development of industrial technologies "Industry 4.0".
This diagnostic system allows to detect failure in the early stage.
The intelligent diagnostic system combines a range of hardware:
-
Speedgoat Performance real time target machine
-
SKF Multilog on-line systems
-
Siemens S7-1200
-
archiving and visualization computer
-
set of sensors - temperature, accelerometers, voltage, currents etc.
The main control software is developed in Matlab / Simulink.
The Digital Twin consists of:
-
two 12-pulse 4Q thyristor converters 6300A, 1050V, connected in parallel, with the ability to work separately,
-
supply network,
-
electrical model of DC motor 7100kW,
-
heat model of DC motor 7100kW with 145 estimated parameters.
12-pulse 4Q thyristor converters in Simulink model
Functions:
DC motor diagnostics:
-
DC motor temperature estimation based on detailed thermal model (145 differential equations)
-
wear rate calculation of the DC motor insulation resource based on the estimated results of the thermal model
-
prediction the time period for bringing the DC motor to overhaul (in developing)
Power electric converter diagnostics:
-
monitoring the status of the converter and determining its faults using a Digital Twin
-
automatic analysis of an emergency situation - determination of the type of failure, determination the number of a faulty thyristor, overvoltage monitoring, localization of the cause of the failure
Mechanical diagnostics:
-
analysis of the state of the mechanical part of the mill stand based on statistical data obtained from vibration sensors of the elements of the stand of mill 1700, signaling about deviation of parameters from the normal mode (in developing)
Measurements in Speedgoat Performance real time machine:
-
40 analog channels with a sample time down to 50 μs (electrical values)
-
via Modbus TCP (Master) - with a sample time down to 1 ms (electrical measurements, accelerometer signals)
-
via Profinet (Slave) with Siemens S7-1200 - with a sample time of 10ms
-
reading a Comtrade-file from recorders for automatic analysis of an emergency situation
Additional system features:
-
the system has a built-in 4G modem for connecting the diagnostic system to the Internet,
-
using a VPN, a network is organized that allows us to remotely perform work on setting up, monitoring, changing the algorithms of all components of the system
Bus signals in Simulink model