A5060026142- Advanced Combustion Engine Technologies
- Process Optimization and Integration
- Biofuel production and bioconversion
- Microbial Metabolic Engineering and Bioproduction
- Computational Drug Discovery Methods
- Energy Load and Power Forecasting
- Biodiesel Production and Applications
- Advanced Control Systems Optimization
- Machine Learning in Materials Science
- Smart Grid Energy Management
- Catalysis for Biomass Conversion
- Heat transfer and supercritical fluids
- Electric Power System Optimization
- Model Reduction and Neural Networks
- Energy Efficiency and Management
- Integrated Energy Systems Optimization
- Thermochemical Biomass Conversion Processes
- Scheduling and Optimization Algorithms
- Particle physics theoretical and experimental studies
- Protein Structure and Dynamics
- Extraction and Separation Processes
- Chemical Thermodynamics and Molecular Structure
- Combustion and flame dynamics
- Fault Detection and Control Systems
- Hybrid Renewable Energy Systems
Forschungszentrum Jülich
1992-2025
Stadtwerke Jülich (Germany)
1993-2019
RWTH Aachen University
2012-2017
Increasing carbon dioxide accumulation in earth’s atmosphere and the depletion of fossil resources pose huge challenges for our society and, particular, all stakeholders transportation sector. The Cluster Excellence ‘Tailor-Made Fuels from Biomass’ at RWTH Aachen University establishes innovative sustainable processes conversion whole plants into molecularly well-defined fuels exhibiting tailored properties low-temperature combustion engine processes, enabling high efficiency low pollutant...
Prediction of combustion-related properties (oxygenated) hydrocarbons is an important and challenging task for which quantitative structure–property relationship (QSPR) models are frequently employed. Recently, a machine learning method, graph neural networks (GNNs), has shown promising results the prediction relationships. GNNs utilize representation molecules, where atoms correspond to nodes bonds edges containing information about molecular structure. More specifically, learn...
The assessment of the ignition quality a wide range oxygenated hydrocarbons is one key challenge in identification novel molecular entities qualifying as biofuels or biofuel blend components derived from oxygen-rich lignocellulosic feedstocks. present contribution summarizes results comprehensive experimental screening campaign targeting diverse set pure-component hydrocarbon fuels and their characteristics an ASTM D6890 Ignition Quality Tester (IQT). This constant-volume combustion chamber...
Oxygenates derived through the selective catalytic refunctionalization of carbohydrates lignocellulosic biomass can be tailored to exhibit desired physicochemical fuel properties that unlock full potential advanced internal combustion engines. Considering fuel's molecular structure a design degree freedom, we present framework for model-based is envisioned guide experimental investigation toward most promising entities. Following generate-and-test approach computer-aided design, novel...
Oxygenated species obtained from the selective chemo-catalytic refunctionalization of lignocellulosic materials and rationally formulated mixtures thereof can be tailored to needs advanced internal combustion engine concepts like low temperature compression-ignition or highly boosted spark-ignition combustion. In present contribution, we a framework for model-based formulation biofuel blends with properties by considering fuel's molecular composition as fundamental design degree freedom. To...
Electrifying energy-intensive processes is currently intensively explored to cut greenhouse gas (GHG) emissions through renewable electricity. Electrification particularly challenging if fossil resources are not only used for energy supply but also as feedstock. Copper production such an process consuming large quantities of fuels both reducing agent and supply. Here, we explore the techno-economic potential Power-to-Hydrogen decarbonize copper production. To determine minimal cost on-site...
Co-design of alternative fuels and future spark-ignition (SI) engines allows very high engine efficiencies to be achieved. To tailor the fuel's molecular structure needs SI with compression ratios, computer-aided design (CAMD) renewable has received considerable attention over past decade. date, CAMD for is typically performed by computationally screening physicochemical properties single molecules against property targets. However, achievable efficiency result combined effect various fuel...
Tailored electrochemical cross-coupling enables a flexible modular access to bio-fuels starting from biomass derived mono- and di-acids renewable energy.
The shift from fossil to renewable fuels presents an opportunity tailor a fuel’s molecular structure and composition the needs of advanced internal combustion engine concepts, while simultaneously aiming for economic sustainable fuel production. We have recently proposed method computer-aided design tailor-made that integrates aspects both product production pathway design. present paper sets out sequentially combine with experimental investigation on single cylinder research model-based...
Demand response (DR) of large industrial electricity consumers is a promising option to balance the fluctuating supply by renewable energies in grid. Renewable energy technologies themselves depend on copper as key material. At same time, production power-intensive process but its DR potential has not yet been quantified detail via scheduling. Here, we analyze optimally scheduling batch and continuous tasks representative process. To determine optimal schedule, formulate mixed-integer linear...
Abstract Fuels with high‐knock resistance enable modern spark‐ignition engines to achieve high efficiency and thus low CO 2 emissions. Identification of molecules desired autoignition properties indicated by a research octane number sensitivity is therefore great practical relevance can be supported computer‐aided molecular design (CAMD). Recent developments in the field graph machine learning (graph‐ML) provide novel, promising tools for CAMD. We propose modular graph‐ML CAMD framework that...
.Metalearning of numerical algorithms for a given task consists the data-driven identification and adaptation an algorithmic structure associated hyperparameters. To limit complexity metalearning problem, neural architectures with certain inductive bias towards favorable structures can, should, be used. We generalize our previously introduced Runge–Kutta network to recursively recurrent superstructure design customized iterative algorithms. In contrast off-the-shelf deep learning approaches,...
This work studies synergies arising from combining industrial demand response and local renewable electricity supply. To this end, we optimize the design of a generation storage system with an integrated scheduling continuous power-intensive production process in multi-stage problem. We both total annualized cost global warming impact consider photovoltaic wind generation, electric battery, trading on day-ahead intraday market. find that installing battery can reduce emissions enable large...
Trading on the day-ahead electricity markets requires accurate information about realization of prices and uncertainty attached to predictions. Deriving forecasting models presents a difficult task due price's non-stationarity resulting from changing market conditions, e.g., changes energy crisis in 2021. We present probabilistic approach for using fully data-driven deep generative model called normalizing flow. Our modeling generates full-day scenarios based conditional features such as...