Welcome to Artificial Intelligence in Electrical Energy Systems
Description: Artificial intelligence (AI) has recently become ubiquitous in our society, with applications in search, image understanding, apps, mapping, medicine, drones, and self-driving cars. Core to many of these applications are visual recognition tasks such as image classification, localization and detection. On the other hand, AI has also been used in Electrical Engineering for many decades for data-driven modelling of system whose analytic modelling was hard or simply not possible to do. For instance, they have been widely used for load/generation/price forecasting, for modelling nonlinear parts of the industrial control systems and for creating surrogate models of complex systems. All these applications are enabled by the artificial neural networks – the fundamental workhorses of the AI.
However, as opposed to past decades when neural networks were small and comprised only a few neurons, recent dawn of the big data age (characterized by the unprecedented access to large computational resources and big datasets) has enabled the creation of much larger networks. They have greatly advanced the computer vision field but have recently also enabled many new applications in the electrical engineering field.
In a nutshell, this course is focused on providing the attendees the following material: a) AI historical background in electrical engineering and wider, b) an understanding and fundamental characteristics of artificial neural networks, and c) practical applications of AI proposed by the lecturers that have solved some of the long standing research problems in electrical engineering. All models will be provided to attendees and experimental lab demonstration is expected as well.
Upon completing the course, the participants will gain insight how to obtain a high accuracy surrogate model of a power electronics systems, utilize the surrogate for multi-objective optimization problems or synthesis of complex power electronics controllers.
Day 1: General information about AI, Tomislav Dragičević (4.5 hours) + Mateja Novak (2.5 hours)
9:00 – 10:00 Artificial intelligence – how is it revolutionizing the world
10:00 – 10:30 Features of neural networks – the workhorses of AI
10:30 – 11:00 Coffee break
11:00 – 12:00 Artificial intelligence in electrical engineering – historical background
12:00 – 13:00 Lunch break
13:00 – 14:00 Application of AI to optimize design of power electronic systems
14:00 – 15:00 Imitation learning of computationally heavy controllers
15:00 – 15:30 Coffee break
15:30 – 16:30 AI for optimizing the control parameters of industrial control systems
Day 2: AI laboratory exercises. Tomislav Dragičević (3.5 hours) + Mateja Novak (3.5 hours)
08:30 – 10:30 Laboratory 1: AI-aided design for reliability of power electronics systems
10:30 – 11:00 Coffee break
11:00 – 12:00 Laboratory 2: AI-aided imitation learning of industrial control systems (part 1)
12:00 – 13:00 Lunch break
13:00 – 14:00 Laboratory 2: AI-aided imitation learning of industrial control systems (part 2)
14:00 – 15:00 Laboratory 3: AI-aided tuning of control parameters (part 1)
15:00 – 15:30 Coffee break
15:30 – 16:30 Laboratory 3: AI-aided tuning of control parameters (part 2)
Prerequisites:
General knowledge about electrical engineering field.
Practicing knowledge in power electronic systems.
Experience in using Matlab/Simulink (Deep Learning toolbox, Parallel Computing toolbox)
The course is recommended for PhD students and power electronic control engineers focusing on multi-objective optimization problems and implementation of complex control algorithms
Form of evaluation: Report evaluated by the lecturers.
Key References:
T. Dragičević, P. Wheeler and F. Blaabjerg, "Artificial Intelligence Aided Automated Design for Reliability of Power Electronic Systems," in IEEE Transactions on Power Electronics, vol. 34, no. 8, pp. 7161-7171, Aug. 2019.
T. Dragičević and M. Novak, "Weighting Factor Design in Model Predictive Control of Power Electronic Converters: An Artificial Neural Network Approach," in IEEE Transactions on Industrial Electronics, vol. 66, no. 11, pp. 8870-8880, Nov. 2019.
M. Novak, T. Dragicevic and F. Blaabjerg, "Weighting factor design based on Artificial Neural Network for Finite Set MPC operated 3L-NPC converter," 2019 IEEE Applied Power Electronics Conference and Exposition (APEC), Anaheim, CA, USA, 2019, pp. 77-82.
S. Mohamed, S. Rovetta, T. D. Do, T. Dragičević and A. A. Z. Diab, "A Neural-Network-Based Model Predictive Control of Three-Phase Inverter With an Output LC Filter," in IEEE Access, vol. 7, pp. 124737-124749, 2019.
Organizer: Prof. Tomislav Dragičević (tdr@et.aau.dk)
Lecturers: Prof. Tomislav Dragičević, dr. Mateja Novak
ECTS: 2
Time: 24 – 25 September 2020
Place:
Zip code:
City:
Number of seats: 30
Deadline: 3 September 2020
Important information concerning PhD courses: We have over some time experienced problems with no-show for both project and general courses. It has now reached a point where we are forced to take action. Therefore, the Doctoral School has decided to introduce a no-show fee of DKK 5,000 for each course where the student does not show up. Cancellations are accepted no later than 2 weeks before start of the course. Registered illness is of course an acceptable reason for not showing up on those days. Furthermore, all courses open for registration approximately four months before start. This can hopefully also provide new students a chance to register for courses during the year. We look forward to your registrations.
- Teacher: Mateja Novak