Dr.-Ing. Peter Bazan

Head of Group Smart Energy

Department of Computer Science
Chair of Computer Science 7 (Computer Networks and Communication Systems)

Room: Room 06.134
Martensstr. 3
91058 Erlangen

Short Biography

Peter Bazan studied Computer Science at the University Erlangen-Nuremberg, Germany and obtained his master’s degree (Dipl.-Inf. Univ.) with a thesis on the area of performance modeling. He worked a couple of years as a consultant for industry projects before he went into business for himself. In addition to his business he worked for the Performance Modelling and Process Control research group at the University Erlangen-Nuremberg. Since 2001 he worked in various positions at the Department of Computer Science (computer networks and communication systems) at the University Erlangen-Nuremberg, where he also received his Ph.D. degree (Dr.-Ing.). He is mainly interested in the modeling and analysis of complex systems. This includes the performance evaluation of computer networks, communication and production systems, as well as the hybrid simulation of renewable energy systems and smart grids.

More Information

  • Hybrid Simulation of Smart Energy Systems

    (Own Funds)

    Term: since 2017-01-01
    The expansion of renewable energy sources and the growing share of decentralized and highly fluctuating energy producers pose complex challenges for modern energy systems. Storage systems such as CHP systems with heat storage, pure electricity storage, or other technologies also play a decisive role. Furthermore, communication between producers, consumers, and storage as well as the intelligent control of electricity producers and consumers is crucial for the stability and efficiency of the energy system.

    The aim of the project is the development of methods and tools for the comprehensive analysis of the increasingly renewable energy-based energy industry at the level of individual houses and smart grids. As part of the project, the simulation framework i7-AnyEnergy is being developed, which enables the rapid development of hybrid simulation models of networked intelligent energy systems. For this purpose, methods such as discrete event simulation (e.g., for consumer, weather, and control models) and system dynamics models (e.g., for energy and cost flows) are combined in a simulation model. The simulation framework i7-AnyEnergy provides basic components for the energy consumption (electrical and thermal), for energy production (eg gas heating, combined heat and power with fuel cells), for renewable energy (photovoltaic), for energy storage (batteries, chemical storage such as based on LOHC), as well as for the control. These components are used to create house models that can be coupled to smart grid models with a common weather model and a communication Network.

  • Dynamic Simulation of Energy Flows and Storage of Waste Heat from Data Centers and of the Integration of Large Storage Systems in Local Heating Networks

    (Third Party Funds Group – Sub project)

    Overall project: Energie Campus Nürnberg 2
    Term: 2017-01-01 - 2019-12-31
    Funding source: Bayerische Staatsministerien
    The share of electricity from photovoltaics in the electricity mix in Germany has been greatly expanded in recent years. In the near future, electricity generation from renewable energies and thus also solar generated electricity will continue to increase. At high solar radiation, this already leads to a local oversupply in the power grid, while the photovoltaic at night naturally can not contribute to the power supply. Ensuring the nightly base load at night is therefore largely ensured by fossil production from coal and lignite with corresponding CO2 emissions.

    By using base load storage systems with low-temperature storage, the use of polluting thermal power plants should be reduced. During the day, heat from geothermal energy or industrial processes is upgraded with excess electricity from photovoltaics using heat pumps (HP) and stored in a low-temperature storage system. To generate nocturnal base load power, this heat energy is then removed from the storage via an Organic Rankine Cycle (ORC) process.

    The aim of the project is the dynamic simulation of energy flows in HP-ORC heat storages that are integrated into the energy system and use excess heat and power. With the simulation models, the dimensioning and suitable operating modes for the economic operation of low-temperature storage systems should be investigated.

  • Modeling and Analysis of Complex Systems

    (Own Funds)

    Term: since 2016-01-01
    Today's computer technology supports researchers and scientists in developing their complex ideas and innovative technologies. The use of such new ideas and technologies in an increasingly complex overall technical and ecological system will be examined in this project. These can be production systems, transport systems, computer networks, smart grids, or even a combination of such systems.

    The modeling and analysis of such complex systems is supported by powerful data structures and algorithms, which enable the use of common PCs for the calculations. For example, data structures such as Multi-Valued Decision Diagrams (MDDs), analytical methods from queuing theory, hybrid simulation, Mixed-Integer Linear Programming (MILP), and combinations tailored to the system are used.

  • Hybrid Simulation of Intelligent Energy Systems

    (Third Party Funds Group – Sub project)

    Overall project: Energie Campus Nürnberg
    Term: 2011-10-01 - 2016-12-31
    Funding source: Bayerische Staatsministerien
    Development of the simulation framework i7-AnyEnergy for networked smart energy systems. For this framework methods such as discrete event simulation (e.g. for demand, weather and control models) and System Dynamics models (e.g. for energy and cost flows) are combined in one simulation model. It provides pre-built components from which more complex energy system models can be constructed in a flexible way. From the basic components for the energy demand (electrical and thermal), for energy conversion (e.g. gas heating, combined heat and power with fuel cells), for renewable energy (photovoltaics), for energy storage (batteries, chemical storages), as well as for the control, house models can be constructed and coupled to larger systems with a common weather model and communication Network.
  • WinPEPSY-QNS - Performance Evaluation and Prediction System for Queueing Networks

    (Own Funds)

    Term: 2004-11-01 - 2005-12-31
    In cooperation with Chair 4 (Distributed Systems and Operating Systems), the queueing network analysis tool WinPEPSY-QNS (Windows Performance Evaluation and Prediction System for Queuing Networks) is being developed. The analytical methods used include, among others, the mean value analysis (MVA) and the method of Marie. For validation and analysis of non-product form networks, a simulation component will be integrated into WinPEPSY-QNS. In addition, the Jackson method is being used to analyze open product form networks and a decomposition method for open non-product form networks. Of particular note is the possibility to present the results in a tabular or graphic form in a very clear manner and to easily carry out extensive experiments for a given queueing network.

    The analysis methods for queueing networks with general distributions developed in the project Analysis Methods for Non-Markovian Models are to be integrated into WinPEPSY-QNS. They allow the approximative analysis of non-product form networks based on the method of supplementary variables or modeling by phase-type distributions and avoid the problem of state space explosion. Therefore, queueing systems with any distribution of service times and inter-arrival times can be analyzed. Emphasis should be placed on heavy-tailed distributions and the deterministic distribution. Thus, it should be possible to investigate the influence of methods and systems for the protection of privacy, authentication, guarantee of integrity as well as the anonymization on the performance (throughput and delay) through queueing systems.

  • Analysis Methods for Non-Markovian Models

    (Own Funds)

    Term: 2001-11-01 - 2004-10-30
    Traditional approaches to solve non-Markovian models use phase-type expansions, apply the method of supplementary variables or construct an embedded Markov chain. All three approaches have also been investigated in the context of queuing networks and stochastic Petri nets. The phase-type expansion approach suffers from the enlargement of the state space, whereas the method of supplementary variables and the embedded Markov chain construction require basically that non-exponentially timed activities are not concurrent. If they are concurrent this will result in multidimensional differential equations that are hard to solve. To avoid these problems more efficient techniques for the performance evaluation of computer networks like web servers or networks of embedded systems have to be developed. In such systems activity durations with large variances (file transfers) as well as deterministic durations (security aspects) arise.

    We have created two new approaches to approximately evaluate the performance models of these systems; the first one is based on the method of supplementary variables and the second one deals with phase-type expansions. We are currently enhancing these approaches and it is planned to combine them for the solution of large non-Markovian models.
    In cooperation with department 4 (Distributed Systems and Operating Systems) the tool WinPEPSY for performance evaluation and prediction of queueing systems was developed. It contains well known analysis methods for open and closed product form and non-product form networks (mean value analysis, Jackson-method, decomposition methods, simulation) and the new state-based analysis methods are integrated.
    In a cooperation with the Telecommunications Laboratory, Communications, Electronics and Information Engineering Division of the National Technical University of Athens simulation models for embedded network processors have been developed. The goal is to enhance the above mentioned methods, so that even performance measures for these large models can be derived.