Author(s): Prakash C. Swain
Linked Author(s):
Keywords: Rtificial neural networks; Streamflow forecasting; River Mahanadi; Rainfall-runoff models
Abstract: One of the most important prerequisites in the operational flood management is the predicted flood value with reasonable lead-time. Advance knowledge of the quantum of floodwater approaching the dam site can be used to route the flood safely through the reservoir, reducing thereby the danger of peak flood flows. The process of flood is basically uncertain and unpredictable owing to its complex and non-linear dependency on a variety of meteorological and topographic parameters. In the real world situation there are numerous external factors, which affect the decision making process. The decision makers always try to avoid risk. In certain critical conditions, they take a decision based on intuition and experience, which cannot be justified with facts and figures. Therefore, it is realized that complex real-world problems require intelligent systems that combine knowledge, techniques, and methodologies from various sources. These intelligent systems are expected to possess humanlike expertise within a specific domain, adapt themselves and learn to do better in changing environments, and explain how they make decisions or take actions. Attempt has been made in the present work, to map this complex phenomenon using Artificial Neural Network (ANN). An artificial Neural Network is characterized by its architecture that represents the pattern of connection between nodes, its method of determining the connection weights and the activation function. It imitates the function of human brain. Real life data are collected from the Mahanadi basin upstream of Hirakud to formulate and validate the model.
Year: 2002