| Title: | Evaluation of Digital Twin based process load optimization – The mine dewatering case |
| Subject: | Embedded systems, Computer science, Distributed systems |
| Level: | Basic, Advanced |
| Description: |
In industry there are many possibilities to optimize process load with respect to electricity cost. This thesis work will focus on one such application, viz. mine dewatering. Generally, some kind of buffer is need to act as “energy storage”. The objective here is to validate possible savings for a longer period based on a realistic scenario (buffer sizes, electricity cost,…). Goals: -Quantify the cost saving benefit of optimized water pumping.
-Verify accuracy of simpler methods to predict optimization potential
A previous MSc thesis at ABB Corporate Research has studied optimization of mine dewatering. A large underground mine typically has multiple pumping stations, with a water reservoir at each level (see figure). The water reservoirs can be utilized to re-schedule pumping to time intervals when electricity prices is lower, and thus reduce the cost of electricity. The previously developed methods need to be improved and tested using real data and actual electricity prices for longer time period (> 6 months).
We are also working on connecting in real-time such optmizations with the game engine Unity 3D for visualization. The work will address the following points: * Calibrate the simulation model with real data from a Swedish mine
* Verify cost saving potential for longer time period (min 6 months)
* Derive a non-optimization based approach to predict savings
* Evaluate prediction method against rigorous optimizations
We look for students showing drive and interest to work with real-world use cases.
The thesis is to be executed within the project D-RODS, where ABB Corporate Research and MDU are partners.
|
| Start date: | 2026-01-12 |
| End date: | 2026-06-15 |
| Prerequisites: |
Academic Background 1. Desired background in control engineering, embedded systems, computer science, or a related field 2. A solid understanding of digital systems engineering principles. 3. Knowledge on mathematical modelling and optimization 4. Experience with programming in MATLAB/Simulink is required. 5. Basic knowledge of machine learning - prior coursework on embedded systems, machine learning / AI. 6. General good skills in programming 7. Proficiency in English, both spoken and written. |
| IDT supervisors: | Muhammad Naeem Ning Xiong |
| Examiner: | Tiberiu Seceleanu |
| Comments: | |
| Company contact: |