Digital Twin of a Continuous Laboratory Process (Tank Cascade)

📖 Overview

This repository contains the MATLAB/Simulink implementation of a Digital Twin (DT) for a laboratory-scale continuous process (a tank cascade model). This project was developed as part of an engineering thesis and is intended to serve as an advanced educational tool for process control classes.

The Digital Twin addresses several key laboratory challenges:

• safety: it allows students to freely test control algorithms (e.g., PID controllers, model predictive control) without the risk of damaging real-world actuators (pumps, valves)
• visualization: it provides insight into unmeasured variables, enabling the estimation and visualization of the process's internal dynamics, thereby deepening the understanding of the phenomena involved
• diagnostics: it serves as a platform for implementing and testing advanced safety interlocks and diagnostic tools.

A key research aspect of this project, included in the repository, is the model parameter identification process. The project provides scripts (identyfikacjaParametrow.m) for implementing and comparing different methodologies:

1. A black-box model (based purely on measurement data)
2. A hybrid grey-box model (which integrates a fundamental mathematical description of the process).

This allows for the analysis of the trade-off between model-fit accuracy and its generalizability and physical interpretability, which is crucial for educational applications.

📂 Repository Structure

The project is organized into the following modules:

01_PLC_control

Contains the source code for the industrial controller (IDEC FC6A) managing the physical rig and communication.

• identyfikacja.pjw: the original ladder logic project file implementing basic logic, scaling, and Modbus TCP/IP communication server.
• Modbus_Register_Map.pdf: Documentation of the memory map used for data exchange between the physical PLC and the Digital Twin.

02_Matlab_Simulink

The core simulation models.

• DT.slx: real-time Digital Twin model with Modbus TCP/IP blocks for bi-directional synchronization with the physical plant.
• DT_offline.slx: standalone simulation model for offline testing and algorithm development.
• macierze_do_modelu.mat: data file containing initialization parameters, state-space matrices, and physical constants required for the simulations.
• main_pid_optimizer.m: PID tuning function using Nelder-Mead algorithm.

03_analysis_scripts

MATLAB scripts demonstrating the mathematical rigor of the project (Identification & Validation).

• synchronize_sigmnal.m: script synchronizing signals - both rising and falling edge ignoring earlier measurements.

04_HMI_visualization

• DT_app.mlapp: a MATLAB App Designer application serving as the Human-Machine Interface. It provides real-time visualization of the process state and allows for control mode switching (Monitor/Control).
• DT_app_offline.mlapp: HMI for the offline model.

⚙️ Requirements

The following software is required to run this project:

• Matlab 2025b (or newer) - for the online version
• Matlab 2024a (or newer) - for the offline version
• Simulink
• Simulink Desktop Real-Time
• Simulink Industrial Communication - for the online version.

🚀 Getting Started

1. Clone the repository to your local machine.
2. Open MATLAB 2024a (or newer).
3. Ensure all required Toolboxes from the Requirements section are installed.
4. Make sure the project folder (including the .mat file) is added to the MATLAB path.
5. To launch the user interface, open and run the DT_app.mlapp application.

🎓 Academic Context

This project serves as the practical implementation of the models and concepts being developed for an upcoming B.Sc. engineering thesis "Digital twin of the continuous laboratory process" at AGH University of Krakow. The model and its identification methods are being verified and validated using measurement data collected from a real-world laboratory rig.