| Acronym | Topic description | Required skills | Contact |
|---|
| DSM_DH | Demand Side Management in District Heating Grids Type of research: Literature review Research question: - What options exist for demand side management in district heating grids?
- What are the pros and cons of the different options?
- What is the potential of demand side management in district heating grids?
Method: - Development of characteristics for evaluation and grouping scientific publications
- Search and read relevant publications on the topic
- Provide summaries in the form of tables and figures
| - Basic understanding of district heating systems
| Daniel Zinsmeister d.zinsmeister@tum.de |
| DHP | Market Agent for District Heating Prosumer
Type of research: Modeling Research question: - How can bids and offers for district heating grid prosumers be generated?
- How can thermal storages influence offers / bids?
Method: - Calculate marginal costs and maximum production capacities of the heat production, based on an existing optimization problem
- Find a method to integrate the flexibility of thermal storages into the pricing scheme.
| - Basic understanding of optimization
- Programming in Python
| Daniel Zinsmeister d.zinsmeister@tum.de |
| EDP | Creating an Energy Data Platform Type of research: Prototyping Research question: - How can energy data be categorized and visualized?
- How can data collection, curation and visualization be done efficiently?
Method: - A: Comparison of existing approaches for data curation
- B: Based on A, development of a small sample data taxonomy (e.g. for renewable generation data only, and for a small number of characteristics)
- C: Implementation of an automated data curation and visualization framework for your sample data taxonomy from B (e.g. automatic recognition of characteristics like time resolution, unit, etc. and automatic generation of time series graphs)
| - Passion for data
- Basic programming skills (or motivation to learn programming in Python within this seminar)
| Annika Schneider
annikakristina.schneider@tum.de
|
| openTUMflex | Flexibility potential of a prosumer-based heating network.
Type of research: modeling Research question: - How much flexibility potential does a prosumer-based heat grid offer to the power sector?
- What influence does the combination of different heat generation types (heat pump, solar thermal, CHP, ...) have on the flexibility offer?
Method: - Create different scenarios in openTUMflex (https://opentumflex.readthedocs.io/) with the components of a prosumer-based heating network.
- Compare the influence of the increasing complexity of the overall system on the result in openTUMflex.
| | Daniel Zinsmeister d.zinsmeister@tum.de |
PSK_modelica PSK_data | Digital twin of a pumped storage power plant model Type of research: modeling Research question: - Create a digital twin for the pumped storage power plant model using several series of measurements.
- a) data driven
- b) with Modelica
- How well does the developed Digital Twin represent the pumped storage power plant model?
- Can any control developed for the digital twin be transferred to the pumped storage power plant model?
Methodological approach:
- Conducting necessary measurements on the pumped storage power plant model.
- Development of the digital twin
- Implementation of a control strategy
| Programming skills a) Interest in machine learning | a) Data driven
Simon Zollner
simon.zollner@tum.de b) Modelica
Daniel Zinsmeister
d.zinsmeister@tum.de |
| NWP_LOCAL | Weather Forecast Selection and Combination Type of research: Data Analysis and Modelling Research question: - How can locally collected weather data be used for better selection or combination of weather forecast models?
Method: - Make a collection of different weather forecast models and collect forecasts from them in a structured fashion
- Collect local weather data
- Evaluate the accuracy of the forecasts with methods from the literature on general and local data
- Prototype a solution which chooses based on the track record of the weather forecasts the best forecasts or a combination of them
- Evaluate your solution
| - Basic programming skills in a high-level language such as Matlab, Python, R, or Julia
- Interest in applied Machine Learning
| Simon Zollner
simon.zollner@tum.de |
| NWP_SOLAR | Literature Review NWP-based Solar Forecasting Type of research: literature review Research question: - What are capabilities, similarities, and differences in publications on NWP-based solar forecasting? (numerical weather predictions)
Method: - Research publications to the topic
- Research metrics and datasets to compare methods
- Evaluate methods
- Compare the results
| - Interest in Data Science
- Interest in Energy Management
- Basic programming skills in a high-level language such as Matlab, Python, R, or Julia are beneficial
| Simon Zollner
simon.zollner@tum.de |
| FC_LOAD | Tutorial Review Electrical Load Forecasting Type of research: prototyping Research question: - What are capabilities, similarities, and differences in of methods for electrical load forecasting?
- How influences the use case setup the best load forecasting methods?
- With how little effort can we reproduce almost as good results as state-of-the-art methods?
Method: - Research publications to the topic
- Research metrics and datasets to compare methods
- Evaluate methods
- Compare the results
| - Interest in Data Science
- Interest in Energy Management
- Basic programming skills in a high-level language such as Matlab, Python, R, or Julia are beneficial
| Simon Zollner
simon.zollner@tum |
OPT_SIM | Optimization of Loop-Based Simulations and Optimization with Julia Type of research: Prototyping Research question: - How can loop-based simulations be made significantly faster while maintaining a high readability
- What are the overarching design principles which can be applied to multiple domains?
Method: - Understand already existing loop-based simulators for e.g. simulation of a (hydrogen) storages in a residential setting
- Understand and apply concepts of e.g., pre-allocations, type stability, loop vectorization, caching, parallelization, multi-threading, static arrays, in the Julia Programming Language
- Perform benchmarks comparing the results and execution time of your modified code vs the original one
| - programming skills in a high level language
| Simon Zollner simon.zollner@tum.de
|
BENCH_FC | Benchmarking of Probabilistic Forecasts Type of research: Literature Review, Prototyping Research question: - How can probabilistic forecasts be benchmarked?
- What are well suited data sets?
- What are well suited benchmark methods for PV forecasting and load forecasting?
- How can be the assessment be fair by, e.g., avoiding peeking into the future?
Method: - Understand state of the art
- Implement models and fit them on datasets
- Evaluate them
- Document code for reproducibility and further research
| - Interest in Data Science
- Programming Skills in languages like Python or Matlab
| Simon Zollner simon.zollner@tum.de
|
STOCH_OPT | Stochastic Optimization in Household Setting with Probabilistic Forecasts Type of research: Prototyping Research question: - How can existing software frameworks be used for stochastic optimization in a household with PV generation and battery storage?
Method: - Understand state of the art in stochastic programming
- Implement deterministic optimization model
- Implement stochastic optimization model
- Research how probabilistic forecasts can be leveraged to improve optimization results
| | Simon Zollner simon.zollner@tum.de
|
HH_SIM | Design of Residential Energy Demand Simulator Type of research: Prototyping Research question: - How to design an accurate simulator of residential energy demand?
- How to model heating/cooling load and include in model?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to simulate residential energy demand
- Evaluate implementation, e.g., by benchmarking results and evaluating computational complexity
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de
|
DEEP_NILM | Deep Learning for Energy Disaggregation Type of research: Prototyping Research question: - How can Deep Learning be applied to Energy Disaggregation / NILM?
- How does Deep Learning perform compared to other methods?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to train DL models on public NILM datasets
- Evaluate performance by benchmarking with methods described in literature
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
DEEP_HEAT | Deep Reinforcement Learning for HVAC Control Type of research: Prototyping Research question: - How can Deep Reinforcement Learning be applied to HVAC Control problems?
- How does Deep Reinforcement Learning perform compared to other methods?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to train DRL models using HVAC model
- Evaluate performance by benchmarking with methods described in literature
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
HYDROGEN_EVAL | Evaluation of Residential Hydrogen Storage Type of research: Data Analysis & Modeling Research question: - How can hydrogen storage be used in a residential setting?
- How does it compare technically and economically to battery energy storage?
Method: - Understand state-of-the-art based on literature
- Develop models to understand technical limitations and economics of hydrogen storage
- Define scenarios and investigate potential of hydrogen storage compared to other technologies, in particular battery energy storage
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
DEEP_WIND | Evaluation of Deep Learning for Wind Power Prediction Type of research: Prototyping Research question: - How can Deep Learning be applied to Wind Power Forecasting?
- How does Deep Learning perform compared to other methods?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to train DL models on public datasets
- Evaluate performance by benchmarking with methods described in literature
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
HP_MODELS | Design and Implementation of Heat Pump Models Type of research: Prototyping Research question: - How can different types of heat pumps be modelled accurately?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to simulate different types of heat pumps
- Evaluate implementation, e.g., by benchmarking results and evaluating computational complexity
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
DIS_STOC_OPT | Distributed Stochastic Optimization Type of research: Literature review Research question: - What can stochastic optimization be used for in energy system operation?
- How can stochastic optimization be distributed to reduce computation time?
Method: - Read and review papers based on well-defined features
- Provide summaries in the form of tables and figures
|
| Christoph Goebel christoph.goebel@tum.de |
BAT_CONTROL | Dynamic Li-Ion Battery Ageing Prediction Type of research: Literature review Research question: - How can the ageing process of Li-Ion batteries be modeled for control applications?
- How can such models be integrated into dynamic controllers?
Method: - Read and review papers based on well-defined features
- Provide summaries in the form of tables and figures
|
| Christoph Goebel christoph.goebel@tum.de |
RC_MODELS | R-C Models for Heat Transfer in Buildings Type of research: Prototyping Research question: - How to design an easy-to-use Python library for fitting R-C-models of the heat transfer in buildings?
- How to use fitted models to simulate buildings in control environments?
Method: - Understand state-of-the-art based on literature
- Develop efficient Python code to fit R-C models
- Evaluate implementation, e.g., by benchmarking results and evaluating computational complexity
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
HONEY_HVAC | Using Honeybee Models for HVAC Control Type of research: Prototyping Research question: - How can Honeybee models be used to develop data-driven HVAC control?
- How well can they be approximated by simpler models for the same purpose?
Method: - Learn to work with basic Honeybee models
- Develop efficient Python code to realize a data-driven HVAC controller
- Benchmark method using basic controller
| - Python programming skills
| Christoph Goebel christoph.goebel@tum.de |
FLOW_NRG | Flow-Based Generators in the energy context Type of research: Literature Review Research question: - For which tasks in the energy context can flow-based generators be applied?
- How can the flow-based generator architecture be exploited/adopted for these specific tasks?
Method: - Learn about flow-based generators and identify their strengths, weaknesses and typical applications
- Find existing studies of flow-based generators in the energy context
- Identify appropriate tasks in the energy context for these networks
- Identify a way to use/adopt these networks for these specific tasks
| - Interest in modern machine learning methods
| Annika Schneider annikakristina.schneider@tum.de
|
|
|
|
|
Template | Template (Copy line and adapt) Type of research: literature review, modelling, experimental, ... Research question: Method: |
| Peter Tzscheutschler |
|
|
|
|
Add Comment