Integrated Prediction Support Tool For WaterTide Predictions

The tasks of modern water management demand accurate and reliable prediction results to support the decision process. The field of water management is mainly supported by highly complex numerical models, which run nearly automatically around the clock. However, there are still many situations, whereas the development of such numerical model systems is too expensive (as well time expensive) or simply not possible because of the complexity of the system to be modelled. With our approach we try to find integration methods for artificial intelligence techniques, such as expert systems and neural networks, to use their advantages without throwing the numerical model systems away. There are several practical reasons to do so. This paper shows a first design of an integrated prediction support tool, which combines traditional model techniques with AI tools by the means of integrated modelling. 1 The tasks of modern water management The tasks of modern water management are highly complex and demand more and more the support of reliable very accurate model predictions Transactions on Information and Communications Technologies vol 20, © 1998 WIT Press, www.witpress.com, ISSN 1743-3517