Table of Contents Author Guidelines Submit a Manuscript
The Scientific World Journal
Volume 2014 (2014), Article ID 432976, 9 pages
Research Article

Accuracy Enhancement for Forecasting Water Levels of Reservoirs and River Streams Using a Multiple-Input-Pattern Fuzzification Approach

1Department of Civil and Structural Engineering, Universiti Kebangsaan Malaysia (UKM), 43000 Bangi, Selangor, Malaysia
2Civil Engineering Department, Faculty of Engineering, Universiti Putra Malaysia (UPM), 43300 Serdang, Selangor, Malaysia

Received 3 December 2013; Accepted 6 February 2014; Published 24 March 2014

Academic Editors: Y.-S. Cho and P. Fuschi

Copyright © 2014 Nariman Valizadeh et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Water level forecasting is an essential topic in water management affecting reservoir operations and decision making. Recently, modern methods utilizing artificial intelligence, fuzzy logic, and combinations of these techniques have been used in hydrological applications because of their considerable ability to map an input-output pattern without requiring prior knowledge of the criteria influencing the forecasting procedure. The artificial neurofuzzy interface system (ANFIS) is one of the most accurate models used in water resource management. Because the membership functions (MFs) possess the characteristics of smoothness and mathematical components, each set of input data is able to yield the best result using a certain type of MF in the ANFIS models. The objective of this study is to define the different ANFIS model by applying different types of MFs for each type of input to forecast the water level in two case studies, the Klang Gates Dam and Rantau Panjang station on the Johor river in Malaysia, to compare the traditional ANFIS model with the new introduced one in two different situations, reservoir and stream, showing the new approach outweigh rather than the traditional one in both case studies. This objective is accomplished by evaluating the model fitness and performance in daily forecasting.