Predictive analysis of unscheduled downtime and optimization of a pipe rolling mill technological processes
METALS
Objective: Predictive analysis of equipment operation to ensure stable quality in continuous production and accelerate decision-making
Tasks:
Solution
Tasks:
- Analyze input and output data of the physical and mathematical model of the rolling mill, process parameters, and data
- Create the architecture of a digital twin of the production system
- Adapt and implement the interface to the system
- Debug and adapt the physical and mathematical model of the pipe rolling mill
- Visualize recommendations for adjusting the technological operating mode of the pipe rolling mill on the platform
- Perform 3D visualization of the hot finishing line
- Build hybrid models and analyze the results of the hybrid models by analyzing scenarios of the hybrid model's operation
- Test the platform with real data
- Develop a user manual for working on the platform
Solution
- Development of diagnostic models for equipment operation to reduce downtime in the rolling mill
- Development of optimization models to minimize the defect rate in the products
- Recommendation of control parameters
- 10-20% reduction in unscheduled equipment downtime
- Early recognition of 54% of all unscheduled downtime events
- 1-5% reduction in the proportion of defective products
- Mill productivity increase by 0.5-2%
Optimization of gas consumption in furnaces
METALS
Objective: Optimization of the gas consumption in heating furnaces for cost savings
Tasks:
Solution
Optimization model based on a neural network that takes into account the interdependencies between control loops to reduce gas consumption
Results
The effects were observed using historical data.
Gas consumption savings:
≈10%
≈ 1.0 million rubles per month
Tasks:
- Optimization of the gas flow control system
- Taking into account interdependencies between control loops
- Incorporation of external system parameters
Solution
Optimization model based on a neural network that takes into account the interdependencies between control loops to reduce gas consumption
Results
- The system generates setpoints for regulators to efficiently utilize the furnace
- Gas consumption reduction in the zones through optimal control
The effects were observed using historical data.
Gas consumption savings:
≈10%
≈ 1.0 million rubles per month
Optimization of steel melting process and reduction of refractory material wear
METALS
Objective: Development of quantitative recommendations for selecting the slag-lining pair to minimize carbon increase in metal within a laboratory experiment
Tasks:
Solution
Tasks:
- Preparation of samples of modeled slags
- Preparation of a cylindrical refractory sample for conducting dynamic investigations of slag resistance
- Refinement of the methodology for determining carbon solubility in industrial slags
- Development of a regression model predicting metal carburization
- Development of a regression model predicting slag carburization
- Development of a regression model predicting refractory wear
- Experimental confirmation of the calculated hypothesis
- Conducting two experiments using slags resulting in the best and worst metal carburization
Solution
- Utilization of a unique combination of physical testing methods and digital modeling
- A platform solution in the form of a mathematical model recommending the optimal slag for a specific type of refractory
- Unique self-developed equipment for refining and validating the refractory wear model
- Forecasting the resource life of refractory products from different suppliers using model/real slags
- Recommendations for the chemical composition of slag to enhance refractory durability while preserving the technological properties of the slag
Effects
Increase in refractory lining durability by at least 5%, resulting in savings of up to $1 million per year
Optimization of steel mixing process in a vacuum degasser (4.3 million tons of steel per year)
METALS
Objective: Optimization of the circulation mixing process of alloying agents
Tasks:
Solution
Tasks:
- Analysis of initial data and formulation of modeling tasks. Creation of a map of required calculation modes. Collection and analysis of initial data
- Development of a mathematical model for the spatial distribution of alloying additives in the filled volume of the ladle and vacuum degasser
- Creation of a calculation model
- Simulation of various technological modes according to the developed map of modes
- Processing, analysis, and summarization of modeling results. Systematization of calculation data
Solution
- Detailed hydrodynamics model of mixing and refractory wear intensity
- Creation of a detailed dynamic model of production processes
- Optimal parameters based on a combination of digital models and real operational data
- Reduction of refractory wear by up to 10%. Average savings of up to $1 million per year
- Doubling the steel mixing speed with ferroalloys
- Increase in productivity by 0.5%
- Increase in production output by $500-600 thousands per year per one vacuum degasser
