Andreas Rummler’s research while affiliated with Technische Universität Ilmenau and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (9)


Fig. 1. Rapid Development Evaluation Platform
RDK - Rapid development kit for mixed-signal systems
  • Conference Paper
  • Full-text available

June 2004

·

143 Reads

·

2 Citations

·

G. Scarbata

·

S. Arlt

·

[...]

·

The goal of evaluation platforms is design time and verification costs reduction, as these are the main parameters affecting time to market and final success of a product. The following article describes a new approach to prototyping of complex mixed-signal systems, suitable especially as a platform for system developers. Both digital and analog subsystems can be emulated within one development kit. The digital part is based on the ARM ASIC-Integrator Kit, targeting rapid development of embedded applications. Patented Electrically Programmable Analog Array (EPAA) acts as the part morphing the functionality of the analog subsystem. EPAA with fully parametrizable architecture allows either for transistor-level design of various analog function blocks, or reuse of pre-designed macros. In-field parameter adjustment and topology configuration designates the EPAA to be used in sensor and actor technology, or signal conditioning. Dedicated configuration software has been developed, allowing for various modi of device configuration, depending on the designer's experience and used abstraction level.

Download

Introducing Ealib - A Java Evolutionary Computation Framework

November 2001

·

54 Reads

·

15 Citations

This article introduces the evolutionary computation framework eaLib. After a short introduction, in which the need for EA toolkits is shown, a generalpurpose genetic representation of individuals is explained. In the next two sections the authors show the concept of breaking up evolutionary algorithms into several independent operators. These operators are transformed into software components with a common interface and used to create algorithms in a fast and flexible way. The papers concludes with a short outlook on future work.


eaLib — A Java Frameword for Implementation of Evolutionary Algorithms

October 2001

·

21 Reads

·

6 Citations

Lecture Notes in Computer Science

This article gives an overview over eaLib, a framework for the implementation of evolutionary algorithms written in Java. After an introduction the kind of genetic representation used in the toolkit is discussed and provided genetic operators are introduced. Thereafter the concept of breaking an evolutionary algorithm into components and the definition of interfaces for these components is discussed. On that basis a controller model for flexibel and fast creation of algorithms is presented. The paper concludes with a section dealing with issues of parallelization of evolutionary algorithms and gives a short outlook on future work.


Structuring Evolutionary Algorithms into Components using the eaLib Framework

August 2001

·

13 Reads

·

1 Citation

Using the terms of evolutionary computation (individual, chromosome, etc.), it is shown that a general-purpose representation of individuals is possible by means of genetic algorithms. The principles necessary in the software to produce genetic algorithms are presented, and the authors then introduce the concept of breaking up evolutionary algorithms into several independent operators and show how to define the interfaces for the operators. A model within which all the operators can function for flexible and fast creation of algorithms is presented and an example is given in illustration. The papers concludes with a short outlook on future work.


MOVE-BASED CROSSOVER FOR GRAPH PARTITIONING PROBLEMS

3 Reads

This paper introduces a specialized crossover operator for use in graph par- titioning problems. The article starts with a short introduction to the graph partitioning problem and its formulation as a multi-objective task. Thereafter a suitable genetic repre- sentation, that is capable of ignoring redundant solutions from the search space is intro- duced. This encoding works together with the proposed crossover operator, that borrows its main idea from conventional move-based partitioning heuristics. The operator has been incorporated into SPEA and has been tested with several benchmark graphs. Experimental results are shown in the last section of the paper.


Evolvica–A Framework for Evolutionary Computation

34 Reads

·

6 Citations

This article gives an overview over Evolvica, an evolution-ary computation framework written in Java. Evolvica features a heavily object oriented approach of breaking evolutionary algorithms into small independent operators. This allows the implementation of a huge variety of genetic algorithm variants. The design of the framework is done in a way to support visual composition of algorithm and flexible enough to extend its structure to distribute the execution of algorithm in a local network. In addition the paper gives a short over framework requirements and already existing toolkits.


Elektrisch-Programmierbares Analoges Array

37 Reads

·

1 Citation

Zusammenfassung Auf Grund der Anforderung von höherer Performance von integrierten gegenüber aus diskreten Elementen aufge-bauten analogen Schaltungen sind neue Ansätze zur Implementierung programmierbarer Bausteine gefragt. FPAAs – Field Programmable Analog Arrays beinhalte ahnlich wie FPGAs einfache, meistens skalierbare, analoge Grun-delemente, die miteinander verbunden werden können. Eine primäre Gruppe möglicher Anwendungen ist die Signal-Aufbereitung in der Sensorik. Dieser Beitrag behandelt einen neuen Ansat uber Implementierung eines elektrisch programmierbaren analogen Chips. Dank der feinkörnigen Struktur ermöglicht er eine sehr detaillierte Anpassung der elektrischen Eigenschaften und durch zusätzliche Blöcke bietet er sich auch für den universellen Einsatz in mixed-signal Systemen an. Messungen an den ersten Test-Chips zeigten eine gutë Ubereinstimmung mit den Simulationsergebnissen, welche eine zuverlässige Charakterisierung ermöglicht.


Fast Mixed-Signal System Prototyping using Unique Programmable Analog Array

13 Reads

Terms like co-design and mixed-signal prototyping have become high attention in the last years, because of the rising need for even shorter time-to-market. Unified solutions for fast design of heterogeneous systems would adequate reduce the overall time needed for their validation as if each subsystem was de-signed separately. In the following, we will present an unique approach to mixed-signal system evalua-tion based on the programmable hardware. The problem of the analog subsystem "on-Chip" prototyping was realized with novel, patented Electrically Programmable Analog Array (EPAA). It was designed for class of sensor signal conditioning problems. EPAA with its fully parameterizable architecture can be configured either at transistor level or using pre-defined function blocks from macro-library. First mea-surements on a couple of test-cases have proven the suitability of our approach, especially for low-power smart sensors' signal conditioning applications.


Citations (5)


... More flexibility and complete AFE functionality was provided by Field-Programmable-Analog-Arrays (FPAA), e.g., relevant FPAA commercial examples [79][80][81][82][83][84] and research examples [85][86][87][88][89][90][91][92][93][94][95], and representative AFE implementations [96][97][98][99][100][101]. ...

Reference:

Integrated Sensor Electronic Front-Ends with Self-X Capabilities
Elektrisch-Programmierbares Analoges Array

... However, these descriptions often contain ambiguous components * Contact via http://stevedower.id.au/ that prevent independent reproduction of experiments without a significant number of correct assumptions. Eiben [14], Peer [29], Ventura [35], Rummler [32] and many others have identified the need for algorithms to be completely shareable and verifiable. Jon Claerbout coined the term " reproducible research " to refer to this need [18] and early 20th century philosophers Cohen and Nagel stated, " scientific method ... is concerned with verification " [4]. ...

Evolvica–A Framework for Evolutionary Computation
  • Citing Article

... EA implementation has been the subject of many works by our group [6][7][8][9][10] and by others [11][12][13][14]. Much effort has been devoted looking for new hardware platforms to run EAs as GPUs [14] of specialized hardware [15]) than trying to maximize the potential of usual hardware. ...

eaLib — A Java Frameword for Implementation of Evolutionary Algorithms
  • Citing Conference Paper
  • October 2001

Lecture Notes in Computer Science

... In-field parameter adjustment and topology configuration designates the EPAA to be used in sensor and actuator technology, or signal conditioning. Additional to [7], we will introduce our EPAA as it could be included in a system for rapid mixed-signal prototyping. We will also present first measurement results of the prototype. ...

RDK - Rapid development kit for mixed-signal systems

... Additionally, we employ representations that only map to a portion of the search space, such as permutation based, in the hope of increasing the probability of finding the optimal solution. All the implementations are based on the Evolutionary Computation Framework [25]. We opted to use multiple genetic operators with every encoding since our previous experience showed such a strategy giving the best results. ...

Introducing Ealib - A Java Evolutionary Computation Framework
  • Citing Article
  • November 2001