Mathematical Modelling
Calculations are based on current engineering methods for gas supply systems and oil/petroleum product transportation in accordance with ISO and GOST standards. Nonlinear system solving via Newton–Raphson and Global Gradient Algorithm (CGA). Real gas equation of state, gas property calculations per trunk pipeline design norms and GERG-2008. Non-isothermal gas flow simulation. Pigging (MFL) simulation. Phase envelope construction and dew point calculation for water and hydrocarbons.
Computing Environment
- Parallel computing: OpenMP, CUDA
- Platforms: Linux, Windows
Design Approaches
The platform is developed using Clean Architecture (layer separation) and Domain-Driven Design (DDD) for modelling pipeline transport domain logic.
Technology Stack
- Backend: C++, C#
- Frontend: Vue.js
- DBMS: PostgreSQL, ClickHouse
- Message queues: RabbitMQ
- Container orchestration: Kubernetes (k8s)
- ML / AI: Python, TensorFlow
Authentication & Authorisation
- LDAP: integration with corporate directories
- SSO: single sign-on
- Keycloak: account, role and token management
Monitoring & Observability
- Prometheus: metrics collection and storage
- Grafana: visualisation and dashboards
Development & CI/CD
- Git for version control.
- CI/CD pipelines configured for automated build, testing and delivery.
Integration with Industrial Systems
- Protocols: OPC UA, OPC DA, OPC AE.
- Operating modes: Off-Line and On-Line (real-time monitoring).
- Integration with SCADA systems.
Machine Learning & Neural Networks
Applied for leak detection, operation mode forecasting and telemetry analysis. Architectures used: CNN (sensor data analysis), LSTM (time series), Transformer (contextual analysis).
New Platform Architecture
Multi-tier architecture, microservices approach, Web UI, REST API, import-independent stack (Apache / MIT / GNU).
Standards & Regulatory Framework
Compliance with ISO, GOST, GERG-2008 and IEC 62541 (OPC UA).