Thomas Preisler

On the combination of top-down and bottom-up methodologies for the design of coordination mechanisms in self-organising systems

In resource-flow systems, e.g. production lines, agents are processing resources by applying capabilities to them in a given order. Such systems benefit from self-organization as they become easier to manage and more robust against failures. In this paper, we demonstrate the conception of a decentralized coordination process for resource-flow systems and its integration into an agent-based software system. This process restores a systems functionality after a failure by propagating information about the error through the system until a fitting agent is found that is able to perform the required function. The mechanism has been designed by combining a top-down design approach for self-organizing resource-flow system and a systemic development framework for the development of decentralized, distributed coordination processes. Using the latter framework, a process is designed and integrated in a system realization that follows the former conceptual model. Evaluations of convergence as well as performance of the mechanism and the required amount of redundancy of the system are performed by simulations.

Self-organizing redistribution of bicycles in a bike-sharing system based on decentralized control

Currently, bike-sharing systems undergo a rapid expansion due to technical improvements in the operation combined with an increased environmental and health awareness of people. When it comes to the acceptance of such systems the reliability is of great importance. It depends heavily on the availability of bicycles at the stations. But, in spite of truck-based redistribution efforts by the operators, stations still tend to become full or empty, especially in rush-hour situations. This paper builds upon an incentive scheme that encourages users to approach nearby stations for renting and returning bikes, thereby redistributing them in a self-organized fashion. A cooperativeness parameter is determined by the fraction of users that respond to an incentive by choosing the proposed stations. It uses a decentralized control process to calculate alternative rent and return stations for each of the stations. These alternatives are then proposed to the users when they approach an empty or full station. The approach is based on a decentralized control framework that allows to equipping different distributed software systems with the control capabilities needed to realize the coordination efforts required to achieve the desired self-organizing properties.

An architecture for a distributed smart grid registry system

The growing dissemination of distributed energy resources with steadily changing power flows in combination with the rising demand for electricity leads to increasing problems in grid management. Planning and managing these resources with information and communication technology as well as automation technology for optimal operational planning while maintaining the systems security and stability is one of the key challenges in the smart grid. In order to enable these volatile resources to offer their capabilities to various actors in open energy markets, a platform for standardized information exchange is necessary. In this paper, we argue that a distributed smart grid registry system will address these needs. An architecture is presented that enables an adaptive system expansion for the growing number of distributed energy resources and actors. Controllers of these energy resources can provide their capabilities, described as standardized energy services to smart grid actors such as aggregators. The system will allow those aggregators to integrate these often fluctuating resources into their business processes. Thus, the planning level of the electricity grid will increase which leads to improved system stability, and better management of the grid constraints.

Energy Service Description for Capabilities of Distributed Energy Resources

The increasing number of volatile Distributed Energy Resources DERs in the electricity grid implies a rising level of complexity and dynamics. The integration and management of these DERs have lead to the introduction of the aggregator role, with the aim of providing energy services to system operators and the market. With regard to the often changing capabilities of DERs, the dynamical aggregation of DERs to meet the demand is still a matter of concern. In this paper a generic description for the capabilities of DERs will be introduced in order to allow the aggregator to efficiently search and find DERs suitable for aggregation. These reduced as possible and abstracted descriptions of the DER capabilities are called Energy Services, which should be complete enough for the aggregators search demands. The Energy Service definition will be part of a recent research project, the Open System for Energy Services OS4ES that is going to enable the aggregator to control dynamically configured large scale Virtual Power Plants with IEC 61850. The results of this project and its field test should contribute to the further development of IEC 61850.

Multi-Agent-based distributed optimization for Demand-Side-Management applications

Dynamic and volatile grid conditions caused by the growing amount of renewable energy producers require the operation of large-scale distributed Demand-Side Management (DSM) applications. This is one of the tasks of the aggregator role in smart grid operation according to the Smart Grid Architecture Model (SGAM). For the optimization of distributed demand-side loads under such conditions, Multi-Agent Systems (MAS) have been shown to provide an appropriate paradigm to model, simulate and deploy automated operating components.

Mesoscopic Stochastic Models for Validating Self-Organizing Multi-Agent Systems

The construction of self-organizing Multi-Agent Systems (MAS) is still short of systematic validation methods. These are crucial for acceptance of self-organization in mainstream software engineering. Validation of such systems requires the use of formal descriptions of the underlying coordinating process during the engineering process. The here proposed approach is based on stochastic validation methods applied to the microscopic states under an periodicity assumption. Thus, we propose in this paper Mesoscopic Stochastic Models as a novel methodological artefact for validating self-organizing multi-agent systems. To demonstrate the power of this method we apply it to a self-healing resource-flow system, in which a decentralized coordination process is used to restore the systems functionality after a failure. Information about the error is propagated through an overlay until the system is able to restore its original functionality.

Simulation as a service: A design approach for large-scale energy network simulations

In the ongoing GEWISS project it is planned to implement a geographical heat information and simulation system. It shall provide a planning and simulation tool for the interlinking of urban development and district heating network development to support the political decision making process in the City of Hamburg. The system shall combine macroscopic and microscopic simulations to a co-simulation system. The simulation as a service approach is presented as a loosely-coupled scalable solution to realize large-scale energy network simulations. It is based on cloud computing technologies for the optimal utilization of computing resources in heterogeneous simulation-infrastructures. This approach can be used to realize simulation systems integrating Multi-Agent System (MAS) based simulations and other simulation technologies. For practical evaluation, two implementation approaches based on a MAS platform as a service-oriented solution will be presented and compared to an approach involving standard web-service technologies.

Towards an agent-based simulation of building stock development for the city of hamburg

In the context of European climate goals municipalities have an increasing interest in an accurate estimation of current and future energy demand in buildings, as the domestic energy consumption is one of the major adjusting screws for the reduction of electrical and thermal energy consumption, whereas the demand for space heating has the highest impact. As part of the ongoing GEWISS project it is planned to create a geographical information system (GIS) to visualize domestic and industrial heat consumption in the city of Hamburg (Germany) to support political decision making by linking the development of urban areas and the district heating grid. Additionally, it is planned to provide simulation capabilities to offer planning assistance for future development. This paper will present the underlying agent-based simulation system that is used to simulate the development of the building stock. Thereby, the simulation approach and first results regarding the development of the renovation state of the building stock based on a study about the renovation behavior of different types of home-owners of detached and terraced houses will be presented.

Dynamic Aggregation of Virtual Power Plants with a Registry System for Distributed Energy Resources.

A growing number of Distributed Energy Resources will be connected to the grid. These distributed energy resources have different technical capabilities, operational constraints and are distributed throughout the electricity network which results in an increasing operational complexity. Aggregators controlling these resources in a virtual power plant might need to reconfigure it in case of failing distributed energy resources (e.g. when the weather changes) or due to changing grid conditions on a system operators request. This means that both, aggregator and system operator, must maintain operational reserves such as frequency control which means also rising costs for the actors. Different approaches such as dynamic virtual power plants have been introduced to address these issues. Extending these approaches, the increasing automation and information as well as communication capabilities will allow the actors to exchange and share real-time data about these resource capabilities on demand, allowing the aggregator to dynamically reconfigure the virtual power plant with external capabilities of distributed energy resources prior operation and during runtime, allowing the virtual power plant to be adaptive towards changing situations. For this purpose a common, active registry system for energy resource capabilities is necessary. In this paper, we will describe a process to dynamically aggregate distributed energy resource capabilities a priori as well as during runtime allowing the adaptive reconfiguration of virtual power plants.

DeCoF: A Decentralized Coordination Framework for Various Multi-Agent Systems

A key requirement to realize modern distributed systems is the ability to adapt the system’s behavior autonomously at runtime towards changing environmental conditions, in order to preserve their operation even in the presence of uncertain changes. The different parts of such a distributed self-organizing system have to be coordinated in order to achieve meaningful adaptations. To avoid single point of failures, decentralized coordination is a key element for the realization of robust and scalable self-adaptation. Due to their inherently decentralized system architecture Multi-Agent Systems (MAS) are well suited to realize such self-organizing systems relying on decentralized coordination. This paper proposes a decentralized coordination framework which focuses on equipping different types of MAS with self-organizing capabilities. Thereby, it shall support various types of MAS so that developers are not limited to a specific platform, while also supporting the coordination of applications consisting out of different, heterogeneous (Multi-Agent) technologies, e.g. required in the area of co-simulations.