http://dl.gi.de/handle/20.500.12116/23847 2026-07-23T22:32:10Z 2026-07-23T22:32:10Z Klinger, Dietmar Page, Bernd Wohlgemuth, Volker http://dl.gi.de/handle/20.500.12116/26591 2019-09-16T09:31:21Z 1999-01-01T00:00:00Z

Realisierung der grafischen Modellierungsmethode Dynamic Relations für zeitreihenbasierte Modelle mit EXTEND Klinger, Dietmar; Page, Bernd; Wohlgemuth, Volker Rautenstrauch, Claus; Schenk, Michael Dieser Beitrag beschreibt die Implementation einer grafischen Modellierungsmethode für zeitreihenbasierte Modelle für die Verkehrsemissionsmodellierung mit dem Simulationswerkzeug EXTEND. Es wird zunächst der Modellierungsansatz vorgestellt. Im Anschluß hieran skizzieren wir die grundsätzlichen Merkmale des offenen Simulationssystems EXTEND. Abschließend wird die Implementation der beschriebenen Modellierungsmethodik mit EXTEND beschrieben und die Vorteile der Verwendung eines offenen Simulationswerkzeuges in diesem Anwendungsbereich aufgezählt.

1999-01-01T00:00:00Z Benn, Wolfgang Beyrich, Günther Görlitz, Otmar http://dl.gi.de/handle/20.500.12116/26592 2019-09-16T09:31:21Z 1999-01-01T00:00:00Z

Semantic Navigation Maps for Information Agents in Environment Information Systems Benn, Wolfgang; Beyrich, Günther; Görlitz, Otmar Rautenstrauch, Claus; Schenk, Michael The automated retrieval of information in Environment Information Systems by information agents is severely hindered by the heterogeneity of these systems. For the decision if an information is a relevant answer to a query, information agents need to understand semantics and context of the query. In this paper we propose our concept of describing and exploiting contextual relations in the domain of environment information on the base of Kohonen’s self-organizing feature maps. We apply this concept for the automated semantic classification of query results into an existing knowledge space. Using this technique, information agents can be endowed with domain knowledge. Thus they are enabled to retrieve information in heterogeneous Environment Information Systems matching the context of the query.

1999-01-01T00:00:00Z Knol, Onno M. http://dl.gi.de/handle/20.500.12116/26590 2019-09-16T09:31:20Z 1999-01-01T00:00:00Z

Developing the CLEAN Model: a Tool to Evaluate Policy Options for Reduction of Mineral surplus, Ammonia Emissions to Air and Nitrogen and Phosphate Emissions to soil Knol, Onno M. Rautenstrauch, Claus; Schenk, Michael Agriculture in the Netherlands produces high emissions of minerals to soil and ammonia to air, which cause environmental problems. The CLEAN model is a tool to evaluate policy options to reduce these problems. An overview of the model is presented and the difficulties in calculating manure surpluses and manure destinations are highlighted. To fulfil its purpose the model needs to be rather detailed, use large amounts of data, be fast, and produce reliable results. The model concepts and information technology that were chosen to meet these requirements are discussed. Special attention is paid to the position of the model in a larger information system and the use of a relational database. The development of this environmental information system will be evaluated and based on this experience, some remarks will be made on model integration.

1999-01-01T00:00:00Z Hasenauer, Hubert Merkl, Dieter http://dl.gi.de/handle/20.500.12116/26588 2019-09-16T09:31:19Z 1999-01-01T00:00:00Z

Mortalitätsschätzungen in ungleichaltrigen Fichtenwäldern mit Hilfe Neuronaler Netze Hasenauer, Hubert; Merkl, Dieter Rautenstrauch, Claus; Schenk, Michael Within forest growth modeling it is understood that individual tree mortality can be captured realistically by relating the average rate of mortality to a few reliable and measurable size or site characteristics using a LOGIT model. In this paper we describe the application of neuronal networks adhering to the unsupervised learning paradigm to predict individual tree mortality. Using the large and representative Norway spruce data sample from the Austrian National Forest Inventory, we train different types of neural network architectures, namely Multi-Layer Perceptron, Cascade Correlation, and Learning Vector Quantization. For training, we use the following learning rules: Error Backpropagation, Resilient Propagation, and Scaled Conjugate Gradient. With an independent data set we evaluate the neural network types to predict individual tree mortality.

1999-01-01T00:00:00Z